The Restructure of Amenities in Beijing's Peripheral Residential Communities

By Meng Ren

Bachelor of Architecture Master of Architecture Tsinghua University, 2011 Tsinghua University, 2013

Submitted to the Department of Urban Studies and Planning in Partial fulfillment of the requirement for the degree of ARGHNE8 Master in City Planning MASSACHUSETTS INSTITUTE OF TECHNOLOLGY at the JUN 29 2015 MASSACHUSETTS INSTITUTE OF TECHNOLOGY LIBRARIES June 2015

C 2015 Meng Ren. All Rights Reserved The author hereby grants to MIT the permission to reproduce and to distribute publicly paper and electronic copies of the thesis document in whole or in part in any medium now known or hereafter created.

Signature redacted Author Department of U an Studies and Planning May 21, 2015 'Signature redacted Certified by Associate Professor Sarah Williams Department of Urban Studies and Planning A Thesis Supervisor Sianature redacted Accepted by V Professor Dennis Frenchman Chair, MCP Committee Department of Urban Studies and Planning

The Restructure of Amenities in Beijing's Peripheral Residential Communities Evaluation of Planning Interventions Using Social Data as a Major Tool in Huilongguan Community

By Meng Ren Submitted in May 21 to the Department of Urban Studies and Planning in Partial fulfillment of the requirement for the degree of Master in City Planning

Abstract

China's rapid urbanization has led to many big metropolises absorbing their fringe rural lands to expand their urban boundaries. Beijing is such a metropolis and in its urban peripheral, an increasing number of communities have emerged that are comprised of monotonous housing projects. However, after the basic residential living requirements are satisfied, many other problems (including lack of amenities, distance between home and workplace which is particularly concerned with long commute time, traffic congestion, and etc.) exist. New remedy plans are undertaken to mitigate such problems. Huilongguan is a typical, representative case of amenity scarcity and improvement dynamic.

The initial aim of this thesis is to investigate whether planning intervention can be evaluated from a crowdsourcing perspective. Using Huilongguan as a case study, research data are obtained from the Huilongguan Community Forum and Dazhongdianping website. The addition of amenities, such as transportation, shopping malls and work zones, is examined to discover how their restructuring affects the daily lives of residents.

Posts on the Huilongguan Community Forum are extracted and categorized according to a specific scheme. The distribution of posts is researched to determine the amenities that have drawn the most attention of residents of this community, as well as the reason. Based on an analysis of sequential content, the attitudes and opinions of residents with regard to amenities are collected and compared to show which amenities are the most satisfactory and those that are problematic. Finally, assessments are made to evaluate amenities from the perspective of users.

Using community forum and public reviews is one approach among numerous others to evaluate planning intervention. In general, traditional evaluations investigate the outcome of planning at a certain time. However, since urbanization is a dynamic process, obtaining prolonged and real-time feedback from different interested parties has become a challenge of traditional evaluations. This thesis is an attempt at using crowdsourced data to evaluate planning intervention. Moreover, this tool provides a more transparent and less time-consuming way to analyze first-hand data in order to assess the outcome of planning. New communities such as Huilongguan are popping up throughout China. Many of these new communities face similar problems as Huilongguan. By evaluating the quality of life in Huilongguan, effective methods can be discovered for decision makers and planners to alleviate such problems.

Thesis Supervisor: Sarah Williams Thesis Readers: Tunney F. Lee Acknowledgement

After the defense, I suddenly notice that this will be the end of the life in MIT and a starting point for another episode. Recalling the time here when I thrilled, struggled, and enjoyed, all the memory become a treasure that I will always appreciate.

I would like to thank all these people for their help and support:

Thank you Sarah and Tunney, for your patient, guidance, and all of the great ideas you give me. I will never be able to go this far without your help. You have set me an example of getting academic achievement with dedication and happiness.

Thank you DUSP community for the caring, inspiration, and fun. I didn't realize how diverse and passion a community can be before I met you folks. I am proud to be one part of it.

Thank you Lei Dong, for your help and inspiration of the scripts. Thank you Zhe Wu, for your tireless help of script and debug. The data grabbing process will be more miserable without your intelligence and expertise.

Thank you my dear friends, Luxi, Fei and Xinyu, for your friendship, endless support and accompany, and all the wonderful discussion about Chinese urban issue.

Thank you my parents and my boyfriend, for all the love and support. It is you who encourage me to pursue my dreams. And it is you who accompany the overstressed nights during the thesis sprint. Words just can't express my affection. Table of Content

Abstract ...... 3 A cknow ledgem ent ...... 4 Table of Content ...... 5 1 Introduction...... 7 1.1 Research Context ...... 7 1.2 Research Questions and Hypotheses ...... 9 1.3 Fram ew ork of Study ...... 10 1.4 Significance and Im plications...... 12 2 Literature Review ...... 13 2.1 Evaluation and Criteria of Planning ...... 13 2.2 The U se of Crowdsourcing in Plan Evaluation ...... 21 2.3 Things That Have Yet to be Explored ...... 22 3 Research Objective and Context...... 24 3.1 Background of Huilongguan's Developm ent ...... 24 3.2 H uilongguan Plan and H istorical evolution...... 27 3.3 Population Dynam ic and Bedroom Com m unity...... 30 3.3.1 D em ographic inform ation ...... 30 3.3.2 Spacial m ism atch ...... 31 3.4 Arising urban problems in Huilongguan: Work-Residence Separation and amenity scarcity ...... 32 3.4.1 Gaps caused by work-residence separation: traffic...... 33 3.4.2 Com m ercial am enity and public services ...... 34 3.4.3 Efforts and im provem ents ...... 34 4 Research M ethod ...... 36 4.1 H istorical and contextual analysis ...... 36 4.2 Interview s ...... 36 4.3 Q uantitative analysis on data...... 37 4.3.1 H uilongguan Com m unity Forum (hlgnet.com ) ...... 37 4.3.2 D azhongdianping (Public Review s)...... 39 5 D evelopm ent of Job Opportunity...... 43 5.1 Introduction...... 43 5.2 Data and m ethods ...... 45 5.2.1 D ata grabbing and processing ...... 45 5.2.2 Content Categorization ...... 47 5.2.3 D ata Analy s ...... 51 5.3 Top D iscussions in the Online Forum ...... 56 5.4 Findings ...... 57 6 Com m ercial Am enities and Services ...... 59 6.1 The Am enity Index ...... 59 6.2 Com parative Evaluation - W ang Jing Area ...... 61 6.3 Food and Beverage ...... 62 6.4 Leisure and Entertainm ent ...... 66 6.5 Sports & Health ...... 67 6.6 Public Services...... 69 6.7 Findings ...... 69 7 Conclusion ...... 72 7.1 Research Results ...... 72 7.2 M ajor findings ...... 73 7.3 Lim itation and Future Research...... 75 7.4 Im p lication s ...... 76 8 B ibliog rap hy ...... 77 A ppendix I ...... 85 The most popular posts regarding to employment bases...... 85 Appendix 11...... 89 The most popular posts regarding to mass transit...... 89 1 Introduction

1.1 Research Context

China's Rapid Urbanization

China's rapid urbanization has led to many big metropolises absorbing their fringe rural lands to expand their urban boundaries. Beijing is such a metropolis and in its urban peripheral, an increasing number of communities have emerged that are comprised of monotonous high-rise apartments and six-story slabs. However, after the basic residential living requirements are satisfied, many other problems exist. These problems include lack of amenities, distance between home and workplace which is particularly concerned with long commute time, traffic congestion, and etc. Huilongguan is a typical, representative case of amenity scarcity and improvement dynamic that involve both planning efforts from top-down mechanism and market-leading individual behavior.

The Huilongguan Case

Huilongguan ([ iRQ) is a peripheral residential neighborhood in northern Beijing. Located in Changping District, Huilongguan is approximately 20 km north from Beijing's city center. Originally a satellite town, Huilongguan has grown to become part of Beijing's extended metropolis inside the 6th Ring Road. After the year 2000, Huilongguan gradually developed into a giant cluster of residential neighborhoods. Its east-west length is approximately 6.2 kilometers while its north-south length is approximately 2 kilometers. During initial planning, Huilongguan's total area was 11.27 square kilometers; its gross floor area was 850 hectares. During the late 1990s, Huilongguan's planned population was 300,000, while its actual population, in early 2000, reached 450,000. Huilongguan Neighborhood has been regarded as Asia's largest community.

7 Figure 1 Huilongguan's photo, scale and location

While rapid urban construction was taking place in Beijing's old city during the 1990s, the large inflow of migrants became associated with the increasing population density. Simultaneously, institutions began to transform with the appearance of a real housing market. Many potential customers were forced into the market despite lacking sufficient economic means to purchase a housing unit. This situation led to massive population relocation outside of central Beijing as well as increased demand for affordable housing.

In order to tackle this issue, in 1998, the Beijing government decided to construct large-scaled residential neighborhoods. These neighborhoods were mainly comprised of affordable housing in Beijing's suburbs. Huilongguan was one of these new towns. In 1998, after the initial planning of the neighborhood by the Municipal Planning Commission, a state-owned developer began to develop the neighborhood. It was developed parallel with the development of Zhongguancun

Science and Technology Park, a Chinese "Silicon Valley." Massive developments emerged, mainly catering to the housing demand of science and education personnel, near the early 2000s.

Huilongguan was labelled as a "Bedroom City" because of its massive population relocation and simplified preliminary planning. For the last decade, the neighborhood has suffered from work-residence separation as well as subsequent problems, such as long commute time and traffic congestion. However, in order to ameliorate the work-resident separation, the municipal government and the local government have been attempting to improve employment allocation and transport capacity near Huilongguan. They also adjusted new commercial amenities in the

8 neighborhood to accommodate local residents' daily life. The problems have been mitigated to some magnitude.

The Huilongguan Community Forum

Huilongguan residents have relatively long commute times to work. The long commutes create compressed time, which in turn causes residents to lose their incentives to establish relationships with others. To some extent, the Huilongguan Community Forum' (hlgnet) seems to facilitate social activities. Since the establishment of the Forum's website, in 2000, it now carries more than 500,000 registered users today. In addition, daily page views are up to 600,000 viewers, which enables the website to be ranked 20th on Chinese Hottest Online Forum site in 20152.

1.2 Research Questions and Hypotheses

Starting with the initial aim of evaluating planning intervention in peripheral neighborhoods of Beijing from a crowdsourcing perspective and using social data, each following aim and study goal will be elevated through the integration of these two research streams. Using Huilongguan as a case study, research data will be obtained from the Huilongguan Community Forum and Dazhongdianping website. The addition of amenities, such as transportation, shopping malls and work zones, will be examined to discover how their restructuring affects the daily lives of residents. This will help determine if crowdsourced data can help city planners evaluate the success of urban development and help guide future projects.

The use of social data as a new tool in this investigation will be crucial. Throughout this thesis, the following three research questions will try to be answered:

1) Can social data be used to evaluate planning strategy and facility quality?

2) How can this tool be used? What are the findings from using this method?

3) What are the limitations and implications of the findings from using this new method? How

http://www.hlgnet.com/ 2 http://www.iwebchoice.com/htmi/class_62.shtml?3Months

9 can planners use this tool to create better plans in the future?

Posts on the Huilongguan Community Forum will be a major source of data. They will be combined with supplementary data on amenities from the Dazhongdianping website. Structured interviews will also be conducted for the purpose of qualitative analysis. These posts will be extracted from the community forum and categorized according to a specific scheme. Then, based on a classification, the distribution of posts on the forum will be researched to determine the amenities that have drawn the most attention of residents of this community, as well as why they have drawn the most attention. Based on an analysis of sequential content, the attitudes and opinions of residents with regard to amenities will be collected and compared to show which amenities are the most satisfactory and those that are problematic. Finally, an assessment will be made to evaluate amenities from the perspective of users.

The following three hypotheses will be examined through the above-mentioned analysis:

1) Social data and content analysis are two of many tools that can be used to evaluate the planning process for amenities.

2) Findings can be extracted from the analysis of social data. Additionally, amenities can be evaluated using this data. These evaluations can reveal the scarcity, partial improvement or complete adequacy of amenities.

3) Summary of steps to sharpen these tools.

1.3 Framework of Study

Before planning is evaluated and a new method is explored, a literature review will be provided in Chapter 2 of this thesis. This literature review will be divided into two aspects, plan evaluation and the application of crowdsourcing. Over the course of plan creation and evaluation, particular criteria should be adopted according to the concepts of various plans and social context (Baer, 1997). The use of crowdsourcing in planning provides collective knowledge and stimulates community empowerment while enabling planners to obtain intelligent and innovative input (Brabham, 2009). After this, the potential integration of planning and crowdsourcing in China will be examined. 10 The general context of the development and dynamic evolution of the amenities of Huilongguan will be explored in Chapter 3. The initial development of this community was rooted in the expansion of the urban area of Beijing and increasing demand for affordable housing. After several rounds of adjustments, a plan for Huilongguan was implemented but the development of amenities was neglected. Following this, specific problems, such as distances between residents and their workplaces, long commute times and the scarcity of commercial amenities, will be discussed in order to sort out planning intervention to mitigate such problems.

Along with exploring the first research question, Chapter 4 demonstrates the framework of methods. Starting with a brief summary of the holistic methodology used for this thesis, detailed information about the context analysis, interviews and quantitative analysis of social data is provided. Specifically, this chapter explores the characteristics of the two sources of data, the Huilongguan Community Forum and Dazhongdianping website, and the reasons why this data is emphasized. The chapter also summarizes the basic steps of data collection and analysis in order to provide a general sense of how these tools are applied.

In addition to the second research question, the following Chapter 5 and Chapter 6, evaluate facilities and amenities, including employment bases and commercial amenities. Data on employment bases and mass transit systems will be extracted from the Huilongguan Community Forum and categorized after being processed. The frequency of keywords (i.e. names of amenities) will then be compared to reveal the focus of netizens and day-to-day influence of amenities. After this, sentiment analysis will be conducted to determine the satisfaction of residents with regard to specific amenities. Representative discussions will then be extracted and summarized to reflect the major concerns of residents in terms of the planning and maintenance of amenities. In Chapter 6, an index that reflects the level of commercial amenities in Huilongguan will be established and a comparative evaluation of Huilongguan and Wangjing, a well-developed neighborhood, will be conducted. Finally, a holistic assessment will be made after comparing various categories in detail.

In Chapter 7, the research results will be determined by highlighting the findings in the three previous chapters. The limitations of this research and potential areas for future research will then be discussed. Finally, the implications of crowdsourcing with regard to better adapting the

11 analysis of social data in order to evaluate planning are highlighted.

1.4 Significance and Implications

Using community forum and public reviews is one approach among numerous others to evaluate planning intervention. In general, traditional evaluations investigate the outcome of planning at a certain time. However, since urbanization is a dynamic process, obtaining prolonged and real-time feedback from different interested parties has become a challenge of traditional evaluations. This thesis is an attempt at using crowdsourced data to evaluate planning intervention. Moreover, this tool provides a more transparent and less time-consuming way to analyze first-hand data in order to assess the outcome of planning.

New communities such as Huilongguan are not only popping up in cities such as Beijing and Shanghai, but also in many second-tier cities in China. Moreover, many of these new communities face similar problems as Huilongguan, including insufficient local amenities and facilities, excessive commute times, traffic congestion, etc. By evaluating the quality of life in Huilongguan, effective methods can be discovered for decision makers and planners to alleviate such problems. Some lessons may also be learned by researching future planning and urbanization.

12 2 Literature Review

This chapter provides a brief review of two related areas, namely, the development and application of planning evaluations, particularly evaluations of amenities and public facilities in the context of urbanization in China, and the use of crowdsourced data in the planning process. However, there is little research that focuses on interdisciplinary areas. This research attempts to integrate these two areas and examine the potential of crowdsourcing in the evaluation of planning in China.

2.1 Evaluation and Criteria of Planning

General Evaluation and Criteria of Planning

How is it decided if a plan is good? What criteria are applied when evaluating a plan? Which tools can be used to assess plans? At first glance, these questions seem to be easy to answer intuitively. However, such questions have long been discussed, but are still difficult to answer. As Berke and French point out:

The quality of a plan is difficult to define. Plannersoften differentiate high-qualityplans from low-quality ones, but are hardpressed to explicitly define the key characteristicsofplan quality. Literatureon planning is surprisinglyscarce when it comes to what constitutes a good plan. The planningprofession has generally avoided this normative question andfocused insteadon the methods andprocesses ofplan creation. (Berke and French, 1994, 237-8)

In the context of planning in the United States, the evaluation of planning can be defined in various ways while several stages of evaluations can be distinguished in the process of creating a plan. These stages are plan assessment, plan testing and evaluation, critique, research and professional evaluation and post hoc plan evaluation. Although similar words are used for all of these stages, identifying who is responsible for an evaluation, when it is conducted and the goals of an evaluation help to distinguish different types of evaluations (Baer, 1997).

The relationships between the various stages of the process and evaluation of planning are demonstrated below:

13 Plan Assessment: Application of criteria for the plan as embodied in the document.

Plan Testing and Evaluation: Choice of best alternative or "the plan" by insider planning team.

Critique Overall plan critique by outside critic, with assorted -even idiosyncratic- criteria.

Research and Professional Evaluation 4., m Concerned with professional improvements in plan, and performed by insiders or outsiders trained as planners and researchers.

Evaluation of reality if no Post Hoc Plan Evaluation, plan had been put in place

Reality Reality Intended Plan Result or Effect (actual result or effect) (actual result or effect If with a plan In place there had been no plan

Figure 2 Various stages for evaluation in the planning process (adapted from Baer 1997)

It is worth noting that post hoc plan evaluation takes place after a plan is adopted and implemented while other stages are performed during the creation of a plan. The purpose of this final stage is to examine if a plan was implemented and, if so, how it performed (Alterman and Hill, 1978) and its effectiveness (Faludi, 1987). However, this raises two questions: 1) when is it appropriate to evaluate the outcome?; and 2) to what alternatives should it be compared? Various permutations of post hoc plan evaluation are described based on comparisons of the real and intended outcomes of a plan, reality and a null case (if there was no plan) and the differences of "unanticipated consequences of a plan" (Baer, 1997). In particular, as a post-modem concept of planning, "dialogue about community" is even more important than physical outcomes. Moreover, "persuasivenessand communicative quality" is an essential plan quality. The

14 involvement of private consultants also has a positive impact on plan quality (Bunnell and Jepson, 2011).

Baer (1997) created lists of partial criteria from literature and synthesized them into a composite list that planning professionals could use as a checklist (Figure 3). Despite this, these criteria are not definitive, but rather advisory and suggestive. Planners should devise their own criteria during the creation of a plan and evaluate potential outcomes.

Adequacy of Context "Rational Model" Consideration Procedural Validity Explain the context and Show basic planning considerations Explain the who and the setting: the what and why based on underlying theoty and its how of the plan-making. of the document. criteria.

Adequacy of Scope Guidance for Implementation Show how the plan is Consider the instruments and the connected to the larger world, agencies and persons responsible for making the plan work.

Data, Approach, Quality of Communication Plan Format & Methodology Clear communication above Other forms of commu- Make clear the technical all else is necessary for a fair nication are found in bases of the plan hearing form others the plan format itself.

Figure 3 Some suggested criteria for planning assessment (adapted from Baer 1997)

Various Focuses of Plan Evaluation

The criteria of planners vary according to distinguished plan concepts.

Appropriate criteriato evaluate a plan are implicit in the concept that the plan embodies. Moreover, in reciprocalfashion, the plan's concept is clarifiedonly by consideringthe criteriaby which it isjudged... Each time a plan is prepared the planners should specify the criteriaby which they expect it to be evaluated; the effort inevitably sharpens their understandingof what they are about. (Baer, 1997)

15 Post hoc evaluation emphasizes various criteria. Some scholars believe plans should facilitate the recreational live of residents. Michalos et al. researched community indicators to measure the effectiveness of recreational programs (Michalos et al., 2011). Numerous topics, such as indicators and standards for "parksand outdoor recreation, leisure and its relationshipto quality-of-life satisfaction, the development of tourism and instruments and sustainability indicatorsformanaging community-based tourism " and other sustainability issues, are evaluated to build detailed models and indicator systems. Sugiyama studied the link between recreational walking and the parameters of open spaces in a neighborhood. The presence of a large, high-quality park within walking distance of one's home may be more important to promoting sufficient amounts of walking for health benefits than its mere presence (Sugiyama, 2010).

Others develop criteria related to sustainability and ecology. The presence of a state planning mandate, mostly applicable to planning in the United States, as well as having a variety of groups participating in the planning process are key factors that increase overall support for the sustainable development principles of a plan (Conroy and Berke, 2004). Ning evaluated affordable housing projects in China based on an ecological housing estate evaluation system. The analytic hierarchy model evaluates plan outcomes from general planning goals to a detailed eco-residence plan (including energy-saving design, eco-green space system, water circulation system, etc.) and 50 detailed indicators (Ning, 2006).

There is a long-held tradition of using quantitative tools to evaluate amenities and assess traffic. Mandl created an algorithm to compute distances and routes in order to evaluate the average transportation costs of passengers who use public transportation in cities (Mandl, 1980). A holistic approach is provided for the development and programming of transportation that includes sets of methodologies to measure economic efficiency, the impact of a project on land use and the environment, etc. (Sinha and Labi, 2011).

Cases of Plan Evaluation in the Development of New Towns

Plan evaluation also varies according to different planning systems and socio-economic contexts, particularly in the development of new towns. Asian scholars who study the evaluation of amenities pay close attention to problems rooted in the culture and social framework of Asia.

16 In Japan, most new towns are large residential estates that are home to workers who commute to a parent city. Rapid economic growth since 1960 has resulted in an enormous concentration of people in the metropolitan regions of Tokyo, Osaka and Nagoya. The supply of public housing has been strained due to a lack of suitable level land, fragmented land ownership and high prices. Houses with reasonable rent are now located great distances from city centers (Kiuchi and Inouchi, 1976).

Because of these circumstances, new towns like Tama were planned and developed. The urban facilities of Tama, Senri and Senpoku, three new towns, have been compared by collecting official statistics that measure the level of public facilities and subjective data on satisfaction from questionnaires. The main problems of Tama include a low-quality living environment and urban facilities and a heavy financial burden on municipalities (Kiuchi and Inouchi, 1976). After 40 years of development, a new analysis of Tama was conducted and revealed that it was shrinking. The research of demographic dynamics reflects the failure of the plan for Tama (Ducom, 2008).

Evaluations of amenities in new towns in Korea show their problems as well. Research by Jun and Hur, which included criteria of commuting costs, showed that the "leap-frog" development of new towns built outside the green belt cost more than the "development of contiguous new towns" (Jun and Hur, 2001). In a study of green space accessibility in new towns in Seoul, an urban green assessment model was adopted. This model showed that the spatial form and arrangement of new towns were more influential in determining accessibility (Eom and Lee, 2009).

Going one step further, many scholars have contributed to evaluating the development of new towns in China. Similar to the situation in Japan, most new towns in China consist of affordable housing. Scholars have dedicated their work to exploring appropriate criteria and standards to guide the creation of plans or performing post hoc evaluations of current land use, amenities and the living quality of residents.

On one hand, research has concentrated on establishing appropriate criteria for built environment. This includes, but is not limited to, criteria for public facilities and amenities (Min Zhao et al.,

17 2002, Zao et al., 2010 and Dongbo Xu et al., 2012), building forms (Pengfei Wu, 2013) and modularization (Yanning Lv, 2012), the configuration of open spaces (Ruojun Lin, 2008 and Yun tan, 2012) and energy-saving technology (Chongbu Gao, 2012, Yuting Liu, 2012).

On the other hand, research has investigated ways to evaluate the level and quality of facilities and amenities and the evaluation results of specific cases using various methods. Since the lack of public facilities and commercial amenities is the most pervasive and serious problem in these new towns, there is a large amount of literature that covers this topic. Some of this literature even analyzes the case of Huilongguan.

Chenglei Li compares Huilongguan, Wangjing and Fangzhuan, three urban fringe communities in Beijing, in terms of the distribution and intensity of retail services (Li, 2006). First, he points out that key features of commercial space are the style, distribution and scale of commerce. He then analyzes the form and problems of the spatial structure of commerce. The groundbreaking data from his study led to new standards to evaluate the level of development of a commercial district, "Commerce Effective Serving Intensity," which give more consideration to the serving radius of different commercial styles.

18 tW------Wur - -- - -_ I - - - -

V7 4/

Z. 7 --- 'En

Figure 4 Retail agglomeration and spacial network (Li, 2006)

Figure 5 Commerce effective serving intensity (Li, 2006)

19 Scholars (Feiran Li et al., 2013) at Peking University led a survey of 550 questionnaires in 2010 on living space in Huilongguan. They evaluated four aspects of living quality, namely, residential space, working space, shopping space and entertainment space. The results showed that, over the previous decade, this large community had played an important role in the formation of residential suburbanization and the decentralization of leisure and shopping space. Suburbanites had a strong ability to deal with the challenges of long-distance commutes between their homes and work places. From this study, planners learned that plans for future peripheral communities should fully consider the connection between work clusters and public transportation sites.

Figure 6 Comparation of Circles of Entertainment Activities in Huilongguan During 10 Years (Li et al., 2013)

Other literature echoes similar problems that Huilongguan faced in its early stages, including distances between home and work and the scarcity of amenities. These problems were reported in the news, which drew widespread attention and triggered extensive discussions (Shijiao Zhu, 2004, Yan Xing, 2005, Guoqiang Zhai, 2007 and Tan Xu, 2009).

Summary

Based on general plan evaluation typology in the United States and cases of plan evaluation in the development of new towns in China during its rapid urbanization, the methods and criteria of plan evaluation are varied. As Baer (1997) argues, planners should establish their own criteria during the creation of plans and evaluate potential outcomes. 20 Most, if not all, cases of evaluation show that a form of "criteria emphasis - methods - evaluation results" has been inherited. Research, particularly that involving Huilongguan, provides evidence of the scarcity of public facilities and commercial amenities in the early stages of the development of a community. However, only a small portion of this research follows up on dynamic outcomes of planning remedies. Furthermore, plenty of time and energy was available to conduct questionnaires (Feiran Li et al., 2013) and connect groundbreaking data (Li, 2006) as part of existing comprehensive analyses. It is necessary to discover new evaluation tools that are less time-consuming and able to track dynamic development.

2.2 The Use of Crowdsourcing in Plan Evaluation

"Crowdsourcing is the process of obtaining needed services, ideas or content by soliciting contributions from a large group of people, particularly from an online community, rather than from traditional employees or suppliers" (Wikipedia). Brabham (2012) points out that widespread Internet access and the increasing use of mobile terminals have made crowdsourcing possible.

Defined as "an online, distributedproblem-solving andproduction model that leverages the collective intelligence of online communities to serve specific organizationalgoals," crowdsourcing must include the following elements: "1) an organizationthat has a task it needs performed; 2) a community (crowd) that is willing to perform the task voluntarily; 3) an online environment that allows the work to take place and the community to interact with the organization;and 4) mutual benefits for the organizationand community. "(Brabham, 2012)

Crowdsourcing has been used in the business, private and public sectors. The amount of literature on the use of crowdsourcing in the public sector, such as policy advocacy, e-government, e-democracy, etc., has increased (Bott and Young 2012). The epidemic of social media has facilitated virtual collaborative environments where netizens interact, share information, organize activities, solve problems, co-create value and improve innovation (Vreede et al., 2013).

Crowdsourcing benefits both organizations and the community because of its unique characteristics. These include: 1) the low cost of sharing and transacting information, which

21 stimulates and inspires active audiences to coordinate activities beyond the existing formal hierarchy (Brabham, 2013); 2) non-expert knowledge is collected, which benefits the planning process in general, and planners are able to break from traditional mindsets and find creative solutions (Van Herzele, 2004); and 3) crowdsourcing encourages crowd wisdom. Surowiecki (2005) points out that "under the right circumstances, groups are remarkably intelligent and often smarter than the smartestpeople within them."

Most uses of crowdsourcing focus on participatory planning. In terms of its transition from for-profit use to public use, a successful crowdsourcing business model can allow citizens to participate in the planning process for projects, thus providing collective knowledge and stimulating community empowerment (Brabham, 2009). Crowdsourcing is used to increase the transparency of the survey process (Evans-Cowley, 2010) when collecting input from the community and facilitates community dialogue. Moreover, crowdsourcing is used to deal with crises. For example, it helps with tracking, reporting and coordinating relief efforts in the event of a natural disaster (e.g. Haiti, Pakistan) or war. Crowdsourcing is being used in various ways within the context of international development.

2.3 Things That Have Yet to be Explored

Countless papers have been written on plan evaluation since the 1980s from different perspectives. However, traditional theories and methods limit themselves in terms of plan outcomes and high time and human costs. On the other hand, the use of crowdsourcing in planning mainly focuses on community planning. Its features, which include a low cost of sharing and transacting information, ability to connect bottom-up input and collective intelligence, show that it has the potential to be used in other areas.

Additionally, the use of crowdsourcing in planning seems to illuminate a method for plan evaluation that provides first-hand, dynamic and more transparent and innovative data with which to evaluate the outcomes of a plan.

Moreover, there is practically zero research related to the rapid urbanization of China. Given the fact that social media is becoming more and more prevalent, the number of new towns (or peripheral neighborhoods) being developed is increasing and civic awareness is rising,

22 crowdsourcing will be a key approach to planning in the future. It is high time to explore the use of crowdsourcing in plan evaluation in China.

Therefore, this research will contribute to interdisciplinary areas of plan evaluation and crowdsourcing in the context of urban areas in China.

23 3 Research Objective and Context

Huilongguan's development is deeply rooted in the development of Beijing's historical and social background in recent decades. Its facilities and amenities, including employment bases, mass transit, and commercial amenities emerged due to its particular social economic and demographic background. Therefore, by understanding so, the planning evaluation, specifically, analysis of residents' opinions regarding amenities can then embed itself in a profound historical and demographical context.

3.1 Background of Huilongguan's Development

The history of the development of Huilongguan is based on the social constructs of Beijing's expansion. The 1990s, Beijing saw a large inflow of migrants, which was associated with an increased population density. The central city suffered a scarcity of available land. At the same time, social institutional transformation started to be conducted, when housing commoditization appeared. This situation caused many residents to settle outside Beijing's centre in order to find affordable housing.

Huilongguan's development is rooted in the following three historical and social contexts.

1. Increasing Demand of Quality Housing

Since the 1980s, the urban population became increasingly concentrated in Beijing's central city because of rapid economic development and improved construction standards. According to research on the recent evolution of Beijing's urban population, in 1998 the total population in central Beijing grew to 2.65 million and reached 2.86 million in 2002 (Jingwen Wang and Qizhi Mao, 2010).

Rapid growth of population and vehicle number in central Beijing had led to excessive density of urban functions. Furthermore, there has been an increasingly heavy burden on urban housing, transportation, amenity, and environment. Consequently, these burdens have resulted in a myriad of urban issues, such as deficient living areas, high housing prices, traffic congestion, air pollution, etc. In the 1990s, in order to optimize land resources, the municipal government began

24 to move populations from central Beijing to suburban areas.

Urban peripheral areas provide advantages due to their lower prices and improved environments. Furthermore, they have become hotspots for housing development (Zhengran Lu, 2007). Some citizens focused their sights on suburban areas for the sake of quality living environments or real estate investments. In 2006, the urban population in central Beijing dropped to 2.07 million, while the suburban population jumped from 4.93 million of 1998 statistic to 7.34 million; this was a growth rate of 56.82%. Among them, Changping's population sharply increased from 0.46 million to 0.83 million, with a growth rate of 80.43% (Lu, 007).

2. Transition of Housing Policy and Strategy

In Socialist China before the 1980s, "housing was treatedas a welfare provision to which everybody was entitled and was ideally intended to be distributedaccording to aformal definition of a minimum requirementand a maximum entitlement of space per person" (Andrusz, 1984: 15). However, in reality housing was not available for such a distribution. Its allocation was actually based on such criteria as occupation, administrative rank,' job performance, loyalty and political connections and was "at the discretionof housing authoritiesat various levels of government or work units, who made the most of their control of this public resource" (Dimaio, 1974; Underhill, 1976).3

During the late 1980s and 19902 China's housing market witnessed a housing commoditization reform and transition. At the beginning of 1988, China's central government held the first National Housing Reform Conference in Beijing. At the conference, the State Council's Housing Reform Steering Group released a document to announce overall housing reform implementation.

In mid-June of 1998, the National Housing Reform and Construction Conference decided to continue the reform of urban and rural housing distribution systems. Additionally, it was decided to build a new multi-level housing supply system consisting mainly of affordable housing. These decisions were made in order to meet the various housing requirements from deferent income

3 ZHOU Min and LOGAN John R. Market Transition and the Commodification of Housing in Urban China

25 groups and accelerated the housing commoditization. In this context, Huilongguan was designed to be one of the initial, affordable pilot housing projects.

3. A New Round of Strategic Adjustment of Urban Spacial Structure

Faced with Beijing's concentric urban sprawl and constantly aggravating urban issues, development and construction of urban peripheral complexes became key elements of Beijing's strategic plan of spacial structure. "Beijing City Master Plan (2004-2020)" schematically adjusted the spacial layout that changed the original single-centre urban sprawl circumstance. This master plan involved building an urban spacial structure of "two axes, two corridors, and multi-centres" and a metropolitan structure of "central city -new town -town." In the plan, central Beijing was planned to be the kernel area, which was embraced by greenbelt and ten peripheral groups.

As an important component of the Qinghe Peripheral Group, Huilongguan was planned to be a large-scaled residential community that also considered commercial services (such as dining and retail), green spaces, and light industry. In accordance with the 2004 master plan, Changping District's government formulated the "Changping Newtown Master Plan (2005-2020)," in which Huilongguan Neighborhood was planned to be

"An urbanperipheral group mainly comprised of affordable housing that provide housing to central city's relocatedpopulation. Its plan should improve the city function with commercial amenity.

26 4

Figure 7 Distribution of fringe groups in the 1992 Master Plan and the 2004 Master Plan

3.2 Huilongguan Plan and Historical evolution

In April 1995, the Beijing Municipal Institute of City planning and Design (BICP) compiled the HuilongguanArea Long-term Plan. This plan schemed to develop a residential neighbourhood whose total land area was 26 square kilometres. This included an isolated green space of 13.2 square kilometres, and the planned construction area was 9.8 square kilometres. According to the plan, the area was bounded by an urban railway ring to the south, Huilongguan South Road to the north, Jingzang Highway to the west, and Huoying Village to the East. Planned population was 15 million (1.53 million / km).

The HuilongguanArea Regulatory Plan (Land Use Plan) of 1997 expanded the construction area and planned population. In 1998, in order to "adapt to the development of affordable housing circumstances and characteristics," the BICP adjusted the Regulatory Plan and partially altered the land use configuration to cut down industrial and infrastructural land use, while increasing residential use and green space. Afterwards, a sequence of detailed site plans were formulated and implemented to guide subsequent development during the time between late 1998 and 2002.

In 1997, following the BICP's initial neighbourhood plan, the Beijing Tianhong Group, a state-owned estate developer, firstly began to develop the neighbourhood. The neighbourhood 27 was planned along with the development of Zhongguancun Science and Technology Park. Near the early 2000s, massive developments emerged when phase one and phase two constructions were launched. Housing was catered to education and science personnel.

Figure 8 Huilongguan's Historical Evolution

The first project phase began in 1997. It included the first batch of affordable housing. In 2000, the second project phase started to be developed. It included affordable housing units of multi-story slabs in a FAR of 1.6 and a sale price of 2600 Yuan/sq. meters. Phase three to phase seven were then successively developed until 2008.

According to the plan, most commercial and public amenities were allocated surround "the Great Ten Mile Boulevard." However, amenity development as well as infrastructure, such as external traffic arteries, surprisingly fell behind residential development. By the end of 2006, when majority of the work in the first two phases of the residential project were accomplished, only part of the commerce in phase one was developed, not to mention the amenity during phase two. Indeed, during that time, the amount of amenities and locations were still under research and adjustment. Meanwhile, several planned parks had construction delays as well. The initial plan didn't arrange any scaled business or industry except for some pre-existing manufacturing estates. Thereby, the neighbourhood, as developed in the initial plan, did not have a live work balance.

(Yin Wang, Director of Detailed Planning Department of Beijing Municipal Institute of City planning and Design)

28

...... Planningof amenity should consider the certain situation of Huilongguan Community. In 1996, the Beijing municipal governmentformulated the strategicplan to change the agriculturalland of North Suburb Farm and adjacentgreenbelt to residentialuse. In 2000, the government executed the, 4th Document of 1997, and calculatedthe current land use using economic equilibrium. It plans to use 10 to 20 years to implement the plan, which is mostly comprised of six-story slabs and smallportion of high-rise apartments. In 1998, the building type was confirmed as affordable housing.

At that time, the Great Ten Mile Boulevard was already constructed, however, without essential undergroundmunicipal pipelines. Consideringthat the street was not conducive to living, it then was planned to be commercialfacilities. land use area was estimated by "nationalstandards of amenity" in the plan process, including residentialuse, industrialuse and commercial use in term of land use plan, but the plan didn't define a certain location of any commercial centre. Nowadays citizens demand of green space, public space and outdoor sport space has shot up. The significance of cultural and recreationalspace in Huilongguan's plan is progressively explicit. We (BICP) suggested to remainingthe land use andplan of Huilongguan'scentral area as well as adjust the land use along the Great Ten Mile Boulevard

In term of internal transportation, a light rail was planned for construction along the Great Ten Mile Boulevard. However, the plan was scrapped when subway line 13 was established. In 2004, the light rail line #13 was constructed and connected Huilongguan neighbourhoods into the network of Beijing's public transit. This linked Huilongguan and the central city, making them more inseparable. Following the city's expansion, in 2005, the new General-planningofBeijing (2004-2020) replaced "satellite town" with "new town" while dovetailed Huilongguan Neighbourhoods with the central city of Beijing.

29 ......

AfttMAS h""

COMMWCW was 0 Pubk3wV"

40. ow *A -A VOW L 37-1

a Ak

Figure 9 (Left) Distribution of Affordable Housing and amenity,

(Right) New construction of Line 13

3.3 Population Dynamic and Bedroom Community

3.3.1 Demographic information

Considering Huilongguan's initial targets, population presents a distinct feature. Residents in the neighbourhoods, generally speaking, are mostly young migrant professionals 4. According to a 2010 survey conducted by researchers in Peking University, 68% of observations were young folks that were less than 32 years old and 64.5% of them have received higher education. Due to China's One Child Policy and age distribution, family size in the neighbourhoods tend to be small: 65% of all observations came from two to three person families. One notable phenomenon is that only 47% of residents in the survey have Beijing Hukou5 , while 62% of populations have Hukou in the greater Beijing area according to 2013 data from the National Bureau of Statistics of China. One specific process of intra-urban migration underlies this particular demographic

4 People work and live in other city other than hometown, but don't have Hukou of the city they currently live. (also see footnote 5 A hukou is a record in the system of household registration required by law in China. A household registration record officially identifies a person as a resident of an area and includes identifying information such as name, parents, spouse, and date of birth. Because of its entrenchment of social strata, especially as between rural and urban residency status, the hukou system is often regarded as a caste system of China. In big metropolis like Beijing, hukou also means the permit to obtain scarce resources for local residence, such as access to vehicle license, permit to buy a house, and qualification of children's primary education.

30 composition.

3.3.2 Spacial mismatch

Table 1 Residents' occupation in Huilongguan

In term of residents' occupation, a large amount of Huilongguan residents work in the IT field or Information technology 17.80% relative fields (Table 1). Though IT jobs tend to Education 7.00% require relatively flexible working locations, the Marketing 5.80% singular planning of Huilongguan Neighborhood Finance & management 4.80% from engagement locations. makes the area distracted Service < 3.00% In the neighborhoods, residents hardly work within Public Officials < 3.00% walkable distance. Media < 3.00%

Comparing geographic distribution of current tucents .).UUYO residents' homes and their job locations in 2003 Retired < 3.00% (Figure 7) proves that locations and engagement locations were somehow fitted. However, in the last decade, only 20% of current residents changed their jobs. Simply put, the majority of residents' living and working place separated after they moved into Huilongguan. Even for the closest IT industrial area, Zhongguancun, the distance is still about 7 miles. For residents who work in the CBD area, the commute distance can exceed 15 miles. This spacial mismatch resulted in severe issues.

31 IT1

IW4--

Figure 10 (left) Current Huilongguan residents' home location in 2003,

(Right) Current residents' job location in 2003 (Li et al., 2013)

3.4 Arising urban problems in Huilongguan: Work-Residence Separation and amenity scarcity

Because of the movement of population to the exurban areas and the preliminary planning with simple urban function, Huilongguan was labelled as a "Bedroom City." As mentioned in previous chapter, Huilongguan was planned as residential area which contains small part of commercial land use. However, since industrial and infrastructural land use had been cut down, the only job opportunity in the neighbourhood is couples of existing manufacturing industry.

(R4) I moved here because old city renewal and relocation. They (the government and developer) asked me to live here, so I live here. Even though here is the Youth EntrepreneurshipPark few people work here. This road is usuallyfull of desolation.

Every day, residents travel dozens of miles from the neighbourhood to central city. This work-residence separation raises subsequent problems such as long commute time and traffic congestion.

Moreover, the work-residence separation isn't the only problems of the community. Amenity scarcity constrains quality life in the neighborhood as well. This problem also comes from the neglect of preliminary planning, when facility and amenity quota was arbitrarily determined

32 without any matched plan of commercial development.

However, the municipal government and the local government have been attempting to improve employment allocation and transport capacity near Huilongguan in order to ameliorate work-resident separation. And they programmed new public facilities within the neighborhood to accommodate daily demand. The problems have been mitigated to some magnitude.

3.4.1 Gaps caused by work-residence separation: traffic

On the one hand, work-residence separation results in large transportation demand. When this demand cannot be satisfied by the existing transportation condition infrastructure in a newly established neighbourhood like Huilongguan, serious traffic congestion emerges.

In Huilongguan neighbourhoods, approximately 300,000 people need to commute via limited couples of arteries and the light rail line #13. Existing roads are not equipped to handle such traffic flows during peak times, while LRL carrying capacity is insufficient as well. It is common to see lines of cars crawling in the main entrance roads and the bumper crowds accumulate in the LRL stations' entrance, waiting to take light rail trains (Figure 11). This traffic congestion also attributes to air pollution and burdens Beijing's already fragile environment.

Figure 11 Traffic congestion in Huilongguan

On the other hand, there is a trade-off between housing price and commute cost that includes money and time (Figure 12). Due to the municipal's subsidy, Beijing's public transit system is cheap. However, the average commute time in Beijing is 1.3 hours, and in Huilongguan, most people's daily commute takes roughly about two or three. . This frustrating time cost

33 tremendously impacts the living quality of Huilongguan residents. As a result, many people choose to move out after their income increases, which the preceding population analysis dynamic in Huilongguan indicates.

W k

Soacial cfstribu*n of w lingness to Day

Wt-er the ty sp aws

F tra sty Are the cal distribt ceWy TuPcrbs ard sub%4,s Figure 12 the spacial distribution of WTP in a city

3.4.2 Commercial amenity and public services

"Nowadays, it is convenient to buy daily necessities. However, I must go to central city for important or deluxe consumption. At ordinary times, my husband and I always work overtime and come back home late. Thereby, we don't have time for shopping in the neighborhood Besides, there is literally no place for shopping." (Zhu, 2004)

An interview ten years ago revealed the scarcity of public facilities and amenities at the early stage of the neighborhood's development. The amenity development fell far behind the rapid construction of housing projects in Huilongguan. Lots of news reported this situation, and even lead to broad discussion of amenity development for affordable housing communities.

3.4.3 Efforts and improvements

During the last 10 years, in order to tackle work-resident separation, the municipal government and the local government have attempted to improve the job and traffic conditions near Huilongguan. First, new job opportunities have been introduced near the neighbourhoods. For example:

34 1 1411-1- - - ___M __ - - __ I .- -I- __ ------__ ------_-MMMMM -2 --- I - , -- __ - -I ______-

These industry parks, at some level, provide new job locations for residents in the neighbourhoods, but the preliminary plan doesn't leave sufficient land stock for commercial and office development, which restricts the improvement for a multi-functional district.

Also, a number of new transportation infrastructure constructed in the past 10 years, which aim to tackle the traffic congestion and reduce the commuting time. For example, according to the Beijing Rail Transport Construction Plan, 3 subways/ LRL run across the neighbourhoods: Line 13 was constructed in 2004, line 8 was constructed in 2012, and Changping line will open in 2015. Also, new BRT lines started running from the CBD to Huilongguan (Figure 13). And, an overpass was built in the most congested area.

Detailed planning interventions will be discussed in the following chapters.

~~4k!

Figure 13 Subway system in Beijing Rail Transport Construction Plan

35 4 Research Method

To provide a comprehensive assessment of the amenities, such as e new construction of light rail lines and addition of employment bases, a mix of several methodologies are employed. This includes qualitative methods such as contextual and historical analysis, in-depth interview, and semi-structured interview, as well as a quantitative analysis of website data. Based on findings using these methodologies, a set of planning principles and recommendations are formulated, which then could be either applied to the development of new towns or to mitigation effort of China's current bedroom communities.

4.1 Historical and contextual analysis

This thesis begins with contextual and historical analyses of the development of Huilongguan Neighbourhood. These analyses track the development from economic and social contexts, the neighbourhood's planning process and evolution, demographic data of the site, existing zoning, amenity allocation, emerged urban problems, as well as local residents' opinion and satisfaction on different amenities.

4.2 Interviews

Interviews were conducted through face-to-face and phone communication, which includes both in-depth interviews and semi-structure interviews. Interviews with city planning experts and professionals last about one hour. One interview was conducted with Yin Wang, the Director of Detailed Planning Department of Beijing Municipal Institute of City planning and Design (BICP). During the late 1990s, Wang's team conducted major efforts of the master plan and land use plan of Huilongguan. The other interview was with a current SPURS Fellow and associate professor from School of Architecture at Tsinghua University, Ying Wang, whose research and teaching concentrate on local community development both in the central city and in fringe areas of Beijing. The focus was put on urban governance of housing development.

The average interview duration with local residents was 30 minutes, varying from 15 to 40 minutes. All interviews were conducted from December 2014 to March 2015.

36 The summary of interviewees is listed as follows.

Table 2 Summary table of interviewees

Planning Residents/ Tenants Professional Name Code Age Gender Occupation housing type range Yin Wang R I 51-56 male security guard staff quarter Ying Wang R2 31-35 female restaurant manager leasehold tenement R3 31-35 female IT engineer commodity R4 26-30 male high school affordable housing R5 26-30 female high school affordable housing R6 19-25 female commercial model leasehold tenement R7 31-35 male small business leasehold tenement R8 31-35 female IT engineer commodity R9 41-50 female restaurant manager leasehold tenement RIO 26-30 male doorkeeper staff quarter R 11 60-70 female retired welfare house R 12 19-25 female housewife leasehold tenement R13 19-25 female college student college dormitory R14 26-30 female housewife leasehold tenement R 15 19-25 female HR manager leasehold tenement R16 26-30 male graduate student college dorm itory R17 19-25 female college student college dormitory R18 36-40 male univerty faculty affordable housing R 19 19-25 male IT engineer affordable housing R20 19-25 male product engineer commodity R21 26-30 male IT engineer leasehold tenement R22 26-30 male IT engineer affordable housing R23 36-40 female housewife leasehold tenement R24 60-70 male retired atTordable housing R25 60-70 male retired affordable housing R26 19-25 male shop assistant leasehold tenement R27 41-50 female restaurant manager commodity R28 41-50 male IT engineer affordable housing

4.3 Quantitative analysis on data

4.3.1 Huilongguan Community Forum (hlgnet.com)

37 Resource Data

Because of long commute time and compression of personal life, residents in Huilongguan lose their incentive to establish relationships with others due to compression of personal time. The Huilongguan Community Forum6 (i.e. hlgnet) seems to promote social activities to some degree. Since 2000, the website has increased to more than 500,000 registered users. In addition, the daily page views are up to 600,000, which enables the website to be ranked 20th in the Chinese Hottest Online Forum site in 20157. According to Alexa's statistics, the community forum's bounce rate ("Percentage of visits to the site that consist of a single page view") reaches 49.60% and daily page views per visitor reaches 8.86, which indicates a relatively high active agreement of its users8 .

The emergence of the community forum reveals the internal desire for favourable neighbourhood relationships and improved environments. Most forum topics are related to residents' daily life. Netizens performance is especially active under topics such as hot news, built-environment proposals, traffic congestion complaints, children's education, inter-league sport games, and dining plans.

(R5) Manyfriends of mine use the hignet. It is a community forum that helps us to get everyday information. For example, I often search and use the group-buying dealsfrom hlgnet, and discuss my child's education situation on it. It really accommodates my life in Huilongguan.

Social capital is the mutually beneficial resources derived from the coordination and cooperation between individuals and social networks. Based on social capital theory, "By analogy with notions ofphysical capitaland human capital--tools and training that enhance individualproductivity-'social capital' refers to features of social organization such as networks, norms, and social trust thatfacilitate coordination and cooperationfor mutual benefit" ( Putnam 1995). Social capital "refersto the collective value of all 'socialnetworks' and the

6 http://www.hignet.com/

7 http://www.iwebchoice.com/html/class 62.shtml?3Months

8 http://www.alexa.com/siteinfo/hlgnet.com#?

38 inclinationsthat arisefrom these networks to do things for each other" (Simon and Schuster, 2000).

Hlgnet has notably affected the accumulation and production of social capital of the entire Huilongguan community, respectively by networks of civic engagement, norms of reciprocity, and trust (Liu, 2005). It develops a new network of civic engagement through catalyzing various entities such as property owners committee, sports clubs, and grassroots charity activities. It also builds the norms of reciprocity by providing a forum for community input, structuring dialogue among different stakeholders and boosting right-preserving campaigns. Furthermore, it establishes real-life community trust through.

Because of The Huilongguan Community Forum is selected as a major resource of data in following analysis, in particular for analysis of job opportunity and massive transit. This is due to the fact that representativeness from hlgnet as well as its positive influences on community social capital building.

In this thesis, the word "post" is used for publishing a forum message and "reply" as a comment message on the aforementioned post. Since the database of hlgnet doesn't tag replies when conducting queries, "message" is used as a general term covering both posts and replies.

Data Collection and Analysis

The data was gathered by submitting queries like "Zhongguancun Life Science Park" and "subway line 8" to the Huilongguan Community Forum search interface. Data was extracted not only from search page results, but also from detailed message content of each post's hyperlink. Then messages are categorized and counted by frequency. A content analysis examines whether such amenities of job opportunities and massive transportation projects a positive affection on local residential life. Detailed analysis will be discussed in following two chapters.

4.3.2 Dazhongdianping (Public Reviews)

Resource Data

Why using Dazhongdianping to analyze commercial amenity instead of Huilongguan

39 - -- , , - -L -- - -.- I -- -MMMMMMMNW- - --- I

Community Forum?

As mentioned before, hlgnet has numerous posts relevant to amenities, such as dinning plans and sports venues. However, when sifting through these kinds of messages, one may surprisingly find the fact that almost all of the messages are trivial and uncategorized. For example, the following post includes a typical group of messages. A netizen said that, "After seeing this piece of news, I am so hungry now!" The post and all eight replies are under this structure:

I am hungry too! only had two buns for dinner. I've already been hesitating whether to by a (r-l) grilled calamaior a pancake for 20 minutes. I think you can have both calamari and pancake. (r-2) It is too late, there won't be any vendors selling pancakes! If you have calamari at (r-3) home, consider making it by yourself! Hahhh! Already had calamari! (r-4) Bad news! Grill vendors may sell calamari with illegal food additive. (r-5) Don't seduce me! I want pancakes too! (r-6)

Ten years ago there is a clean and delicious pancake vendor in Anding Bridge. Then it was (r-7) prohibited soon. So sad. Foreign countries have legitimate food trucks. Why they are not allowed in China? (r-8)

Figure 14 Structure of an online message example

First, because messages on hlgnet lack any direct geographic information like Facebook, and netizens' activities may be distributed throughout the city, it is difficult to decide whether the activity mentioned in the message occurred within the neighbourhood. In the above example, only r-1, r-3, and r-7 contain specific geographical information, which indicates that there are temporary vendors selling pancakes near a Huilongguan resident's home. However, r-4 and r-5 are not informative enough to infer netizens' opinion regarding a specific vendor in Huilongguan.

Second, the data is uncategorized while names of commercial amenities are heterogonous. For example, as mentioned before, Huilongguan had been notorious for its scarce commercial amenities, such as restaurants. However, without a categorized structure, information spreading so widely that a single query of "restaurants" retrieves limited information, while messages only

40 containing restaurants' particular names are neglected like the above example.

Third, even if all useful messages are collected, the information they carry is too trivial. The inference regarding the grill and pancake vendor lacks evidence, such as whether the restaurant's service is sufficient to support the lives of residents or whether netizens hold positive attitude toward the restaurant. Furthermore, this kind of analytic process is a multi-disciplinary problem that may involve machine learning or natural language processing, which falls beyond this thesis' research scope.

Therefore, new data resources that have both identifiable geographic information as well as categorized structure are searched for. Dazhongdianping is selected as an alternative data resource to conduct commercial amenity analysis.

Cuonen COxgm Rosr T*sw* 4~twp card 0tv ~o. 0nm. ....

nSgM* Fanc, So(W. D~ C4h Or

se.. e so.ef KOM* OCOO

Car 4umds d41 t HUMIn ufn Metro LMn 6) 60.70,922 Hi bPWOw.t so ubfta

q, mpW (WUllnggu, ahp) f US* 0000 82 Co'nmwts Hurlcti Wggu"n 107 WeWt Manl Siret, st Fxr, Quitte. in tk~x~qW shop) 5990122 1101W"Noet a*

By on a, &-a WaSi" then for 0*l 76 yuian, 100 yun wortr & fxh is assured

Hadiar, DnIt"

)Ckn*"e t:n, a ro ~ S240rt CM=

s Fbgrb Theasesr t ohn 3Mt Acet orf Yu&*DWPQaon gian

Figure 15 The searching interface of Dazhongdianping

41 Data Collection and Analysis

Dazhongdianping9 ("Public reviews" as its English translation) is a Yelp-like crowd-sourced review network about local businesses in the Chinese market. The website was established in 2003. By the end of 2014, it already covered 300 cities in China, with service registered for over 4 million users. In metropolises like Beijing, this website provides very convincing coverage.

DZDP's search interface allows users to search and retrieve business information according to service category and location. The site's front page provides 13 types of services (Food, Movies, Entertainment, Hotels, Tourism, Beauty, Bridal, Parenting, Shopping, Sports, Daily Service, House and Home, and Automobile). Each category includes different subcategories. For instance, in the "Food" category lists 25 different food genres. A specific area can be searched and the website will filter out the services that are accessible in this area. For every query, the website returns a number of accessible services in different category. For example, searching "Food" in Huilongguan Neighbourhood, the website directs the user to a webpage showing the number of each "food" genre that is accessible in Huilongguan.

By calculating amenity amount and distribution in various genres, this thesis generates a general sense of amenity level among wide range of amenities from restraints to recreational venues. In order to make comparisons to analogous case neighbourhoods, an index of amenity is applied to normalize it. Along with input from interviews of residents, the evaluation of current commercial amenity is concluded. Detailed analysis will be discussed in Chapter Seven.

9 http://www.dianping.com/

42 5 Development of Job Opportunity

5.1 Introduction

During the last 10 years, the municipal government and the local government attempted to improve the job condition near Huilongguan in order to tackle work-resident separation and to boost local economy. New job opportunities have been introduced in and around the neighbourhoods; majority of these employment bases are high-tech R&D campuses that rely on industrial agglomeration and expansion of Zhongguancun Science Park and Universities (Tsinghua University, Peking University, Beihang University and etc.). Their basic information is listed as follows.

Figure 16 Job Opportunity distribution near Huilongguan

43 101. The Zhongguancun Life Science Park

Was planned to be a high-tech campus that is mainly comprised of research and development of life sciences, business incubators, testing and producing, product appraisal, biotechnological project publication, venture investment, international communication, and personnel training. o Phase #1: Construction: October 2011; Land area: 130.5 hectares. o Phase #2: Under construction; Land area: 122 hectares.

102. Shangdi IT Industry Base

Is an integrated, high-tech industrial park led by the electronic information industry. Today, it is a fully equipped enterprise gathered campus that contains research and development, production, management, training, and services. In order to promote the industrialization of high-tech achievements and enterprises' internationalization, the base focuses on the development of computer science, communications, mechatronics, biotechnology, and other high-tech technology. o Southern District: Construction: 2000; Land area: 181 hectares. o North District: Construction: 2004; Land area: 51.1 hectares.

104. Beijing International Information Industry Base

In virtue of the proximity of Zhongguancun Science Park, this project focuses on attracting the world's top 500 enterprises; this includes well-known domestic and foreign enterprises located in the park. The information industry is the leading industry in the park. It includes software, electronic products, with synergetic development of environmental protection, biomedical, and other high-tech enterprises. o Phase #1: Construction: 2008; Land area: 63 hectares. o Phase #2: Under construction; Land area: 79.2 hectares.

44 121. Er'bozi Industry Park

planned to be demolished in future (timeline: unknown)

131. Sanyi Industry Park

The industrial park's products include construction machinery, piling machinery, road construction machinery, cranes, and 25 other categories. In particular, the leading product is the rotary drilling rig, which is ranked first in China's market share. The park is currently China's largest spin drilling rig manufacturing base.

Construction time: 2012; land area: 34 hectares.

132. Hongfu Industry Park

It integrated industrial incubators and residential communities. Construction time: 1996.

These industrial parks, at some level, provide new workplaces for residents in the neighbourhoods, as well as facilitate development of third industrial businesses by providing potential consumers. However, the preliminary plan doesn't leave sufficient land stock for commercial and office development, which restricts the improvement for a multi-functional district.

5.2 Data and methods

5.2.1 Data grabbing and processing

First, a set of queries are submitted using all science parks and R&D campus names as keywords in the community forum's search interface. This returns the user to the search result page, with several lists of topics posted, user names, reply amounts, view amounts, and publication times. Because the time spectrum of such science parks and R&D campuses spans from the early 2000 until now, all data since the forum's establishment are retrieved from queries. In total 1991 messages are returned as query result.

Following is the summary of all query keywords and amount of returned messages.

45 Table 3 Summary of query keywords and returned messages

Workplac Keywor number e ID Name d ID Query keyword keyword_transi of results 101 Zhongguancun Life Science 10101 $ YWiA41 Zhongguancun Life Science 43 Park li Park 101 Zhongguancun Life Science 10102 1 r-- Th4 Life Science Park 129 101 Zhongguancun Life Science 10103 1iAI4 Life Park 286 102 Shangdi IT Industry Base 10201 1 t f t V - Shangdi IT Industry Base 19 102 Shangdi IT Industry Base 10202 iLOiiJL Shangdi Base 3 1 103 Zhongguancun Changping 10301 Hr AU TEid Zhongguancun Changping 14 Science Park Science Park 104 Beijing International 10401 I HkiXfs rI Beijing International 47 Information Industry Base 'II Information Industry Base 104 Beijing International 10402 Lt,] F [r Beijing International 97 Information Industry Base Information Industry Park 104 Beijing International 10403 f Q1% Information Park 629 Information Industry Base 120 Changping Industrial Park 12001 [ 'J'T kI4 Changping Industrial Park 68 121 Erbozi Industry Park 12101 T- Erbozi Industry Park 83 131 Sany Industrial Park 13101 Sany Industrial Park 161 131 Sany industrial Park 13102 [A flI Sany Industrial Park 66 13 I Sany Industrial Park 13103 Sany Technology Park 81 131 Sany Industrial Park 13104 - f F Sany Heavy Industry 171 132 Beijing Hongfu Industry Park 13201 )i W 'I: V-ri Beijing Hongfu Industry 24 l'Ib l H Beijing youth 35 133 Beijing youth entrepreneurship 10501 T demonstration Park entrepreneurship demonstration Park 134 Xiaotangshan Industry Park 13301 LI'.dV4d Xiaotangshan Industry Park 7 Total 1991

For each of these retrieved forum messages, basic information is extracted including hyperlinks, topic, user's name, publication time, reply amounts, and view amounts. While the content of each topic isn't shown in search results, a small script was used to obtain the content of each post by entering its URL. All 1991 results are then merged with identical query keyword tags, such as 10301, to indicate where the results originated.

Since some forum messages contain more than one keyword in their content, such duplicated posts are merged, in which the amount of replies and amount of views are replaced with the sum of the duplicated ones.

46 Subsequently, since the forum uses a word segmentation algorithm'0 , searching for keywords is designed based on string matching instead of machine learning; the returned results include an amount of irrelevant messages. For instance, when searching "Sany," which is comprised of two Chinese characters "san" and "yi,' some results emerge because they have those two characters, but in distinct sentences or even distinct paragraphs. Furthermore, one identical keyword may indicate another object other than the employment base's name. For example, "Sany Heavy Industry" also represents the name of a specific stock, so search results may consist of stock fluctuation. Thereby, after preparatory content analysis, forum messages whose content is actually irrelevant to the query keywords were rejected. 375 results were left for further analysis.

5.2.2 Content Categorization

First question is to examine what types of messages residents post in the community forum in response to the aforementioned employment bases. Also, how these types of posts reflect the influence on residents' real life as well as residents' attitudes towards them is examined.

To investigate the fact, I developed the classification scheme and categorized the sample set. The classification scheme follows a mixed process: topics of sampled messages are identified in a bottom-up manner" and also adopted categorization ideas from existing literature in a top-down manner. Yang et al (2011), Qu et al (2009) and Vieweg et al (2010) investigate how information regarding natural disasters flows in micro-blogging systems and online forums. In a response to major disasters like an earthquake, because of the disaster's proximity and the 140 characters limitation of micro-blogging, the classification of content shows the nature of homogeneity in this circumstance. Of the majority of micro-blogging messages fall into four categories: informational messages, action-related, opinion-related, and emotional related (Yang et al 2011). However, since information in the online-forum discusses relatively long-term events, and since post content has no limitation on text length, the content could be more comprehensive.

Therefore, a classification scheme is developed based on the both topic and sentiment. Following

'0 Word segmentation is the process of parsing concatenated text (i.e. text that contains no spaces or other word separators) to infer where word breaks exist. " The bottom-up classification means researchees define the classification scheme after skimmed through the samples massages.

47 is the summery of the category.

Table 4 Summary of categorization

Category Discription Peremeters

Reprint of news articles a forum message includes reprint of news and author's/repliers' comments supporting support development of employment bases attitude criticizing criticize problems of employment bases neutral other comments that is neutral a forum message that is propose some planning Motions and Proposals action relevant to employment bases which Opinion- may ameliorate existing problems attitude bases related suggesting support development of employment criticizing criticize problems of employment bases describe experience and observation relevant to Narrative of daily experience employment bases

positive attitude show positive attitude towards employment bases satisfaction negative attitude show negative attitude towards employment bases neutral attitude no explicit attitude is held frequency of Informative Messages providing useful information relevant to amenity's name employment bases occurs in the Information- content related frequency of Advertisements ads with a address in employment bases amenity's name occurs in the content

Reprint of News and Articles

In this category, most text embodies the users' attitudes and opinions for or against the event in an article.

For example, a post topic "Sany Headquarters Will Station in Huilongguan Area" has the following content:

A Hundreds of billions-value Company will station, which will benefit Huilongguan's businesses (Reprintedfrom People's Daily online).

China's largest and the world's sixth largest constructionmachinery manufacturerSany will move its headquarterfrom Hunan to Bejing in two months. A Sany official disclosure

48 yesterday, Sany Group' administrativeheadquarter and its corefunctions of business will be fully moved to Sany IndustrialPark in ChangpingDistrict. The official said, Sany determined to move to Beijing, mainlyfor the needs of internationalbusiness development...1 2

This post has 12 replies and 1813 views: o "My first instinct, there will be a lot of outsiders migratingfrom Changsha to Bejing. (smile face)" o "Sany only recruitspretty Hunan girls." o "Is the location near Er'bozi?" o "Congratulations.House value will increase. Propertyowners who have two units or more laugh in your beard." o "What is it?" o "Sany bought a land near Yongwang ten years ago at a price of 35000 Yuan per acre. o "Sany hadfactory buildings there originally." o "It already has brand there. This time, it is the headquartersthat is moving. Furthermore, the area is big enough to build dormitoriesfor ordinary workers migrantfrom Changsha." o "Ha what a good news. Many local workers must be reluctant to move here. Then Sany will recruita considerableamount of workersfrom Huilongguan, which provide lots ofjob opportunity. By the way, Sany prefers Hunan people." o "Are you sure it is not the Sany Industry Park in ChangpingLiu Village?" o "It is good news." o "Does Sany recruit?"

Three replies are positive that the new development of the Sany Industry Park will supply considerable job opportunities to local Huilongguan residents. Three indicate that the park will spur local economic by introducing potential customers to local business and housing markets. Other comments are general neutral discussions about the details. Then, the supportive to unsupportive attitude ratio is 6/0.

12 http://bbs.hlgnet.com/info/u1_24187775/

49 Totally 94 forum messages, approximately one fourth of the results fall in this category.

Narrative of Daily Experience

This category contains posts that describe different employment bases. This includes discovering a beautiful landscape the Zhongguancun Life Science Park. Additionally, posts described the traffic around the Er'bozi Industry Park as consistently awful. In these kinds of posts, posters, and replies usually include subjective commentary on the existence of employment base. Analogously, the positive/negative description ratio is counted. 56 forum messages fall into this category. 57 forum messages fall into this category.

Motions and Proposals

The active community forum somehow facilitates public participation in the planning process and city management. Every year, grassroots CPPCC (Chinese People's Political Consultative Conference) members and local PCC (People's Congress Committee) members post their draft proposals online for suggestions. Users spontaneously post their opinions and views on CPPCC/ PCC motions as well. This category of posts reveals users attitude as for or against as well as echoes solid and detailed proposals that may ameliorate existing problems. Only ten forum messages falls into this category, but most of them are the highest forum posts that shares most amount of views and replies.

Informative Messages

Enthusiastic users summarize useful information, such as a shortcut to Shangdi IT Industry Base during peak hours and a collection of all useful phone numbers in Huilongguan. In this category, another criterion is applied to depict the impact of new development of industry parks. By counting the frequency of keywords, more frequently used keywords imply more commonly used one. 112 forum messages falls in this category, which reaches 28.6% of the entire results.

Advertisements

Most advertisement posts are affiliated with an address that is either the venue of an even or the address of a company, which laterally reflect the degree of involvement of such work places.

50 Therefore, the more frequently an employment base's name appears, the higher possibility the workplace influences people's lives.

5.2.3 Data Analysis

After organizing the results, all employment bases forum messages and their parameters are counted respectively in five categories. The following table shows the summary of the analytic data.

Table 5 Summary of Sentiment Data

Workplace Category Message Supporting Criticizing ID Count I reprint of news and articles 94 124 83 101 Zhongguancun Life Science Park 40 54 58 102 Shangdi IT Industry Base 11 17 11 103 Zhongguancun Changping Science Park 2 3 2 104 Beijing International Information Industry Base 4 4 5 120 Changping Industrial Park 25 40 16 121 Erbozi Industry Park 0 0 0 131 Sany Industrial Park 29 36 13 132 Hongfu Industrial Park 2 3 2 134 Beijing International Information Industry Base I 1 0 2 Narritive of daily experience 57 48 23 101 Zhongguancun Life Science Park 37 33 18 102 Shangdi IT Industry Base 4 7 2 103 Zhongguancun Changping Science Park 2 3 0 104 Beijing International Information Industry Base 0 0 0 120 Changping Industrial Park 0 0 0 121 Erbozi Industry Park 8 6 3 131 Sany Industrial Park 9 6 2 132 Hongfu Industrial Park 0 0 0 134 Beijing International Information Industry Base 1 1 0

Message Positive Negative Count Attitude Attitude 3 Motion and proposals 10 27 3 101 Zhongguancun Life Science Park 4 9 0 102 Shangdi IT Industry Base 2 5 0 103 Zhongguancun Changping Science Park 0 0 0 104 Beijing International Information Industry Base 1 2 0 120 Changping Industrial Park 2 13 2 121 Erbozi Industry Park 4 7 1

51 131 Sany Industrial Park 2 4 0 132 Hongfu Industrial Park 0 0 0 134 Beijing International Information Industry Base 2 12 2

Message Frequency Count 4 Informative messages 112 125 101 Zhongguancun Life Science Park 36 39 102 Shangdi IT Industry Base 11 13 103 Zhongguancun Changping Science Park 12 14 104 Beijing International Information Industry Base 8 8 120 Changping Industrial Park 6 8 121 Erbozi Industry Park 30 35 131 Sany Industrial Park 14 16 132 Hongfu Industrial Park 1 1 134 Beijing International Information Industry Base 13 15 5 advertisements 42 42 101 Zhongguancun Life Science Park 15 15 102 Shangdi IT Industry Base 10 9 103 Zhongguancun Changping Science Park 11 11 104 Beijing International Information Industry Base 0 0 120 Changping Industrial Park 2 2 121 Erbozi Industry Park 5 5 131 Sany Industrial Park 4 4 132 Hongfu Industrial Park 0 0 134 Beijing International Information Industry Base 8 8

Daily influence:

The following chart and table demonstrates the overall distribution of forum message topics across five categories. As mentioned before, this count reflects employment bases influence on netizens daily life.

52 40 35 30 25 20 Is 10

i~.w I~a.~ Iw~wue~m Indwnal ia ~ %amu klai n~ t*uv% 14i) 0j".M it4 Fig6re 17ig F mesA

0 k4* MW 0 4 0VA

10 0 0I

4ot7~iH4W II4t* q

Figure 17 Forum message distribution

First of all, netizens attention and concern for different employment bases vary. The minimum amount of messages about one particular employment base is 3 while the maximum is 132. While, in each forum message, the amount mean of views is 854.2, and amount mean of replies is 3.47. This means each topic regarding employment bases draws attention from 854 users on average, in which an average nearly 3.5 netizens are involved in the discussion of this topic. The median view is 382 and median reply is 1, which means that the distribution of attention is positively skewed. In addition, several topics are highly popular with thousands of views and dozens of replies.

Second, based solely on the above chart, the message topics are markedly concentrated in a couple of employment bases. Zhongguancun Life Science Park ranks first in both total amount of forum messages and in each individual category. Sany Industrial Park, Erbozi Industry Park, and Changping Industrial Park respectively rank in second, third, and fifth with a peak in specific categories: reprint of news and informative one. Shangdi IT Industry Base is ranked fourth, however, distributed more evenly in all five categories. Other employments have less influence on netizens' daily life. 53

---- -MM - Table 6 Relation among industry parks' area, distance and their daily influence indicator

Area/ Distance/ Opinion- Frequency hector mile oriented Zhongguancun Life Science Park 252 4 81 51 Shangdi IT Industry Base 232 3.9 17 21 Zhongguancun Changping Science 65 7.5 4 23 Beijing International Information 155 3.6 5 8 Industry Base Changping Industrial Park 120 7.9 27 8 Erbozi Industry Park 20 2.75 12 35 Sany Industrial Park 34 2.66 40 18 Hongfu Industrial Park 55 2.2 2 1 Beijing II Industry Base 5 0.3 4 21

Third, a positive correlation seems to appear among post amount and industry parks' scale and proximity. Industry parks with larger area and shorter distance tend to be more influential. However, since the sample size is too small (10), regression or other statistic tools cannot be applied to give a quantitative model. Also, longer history of development seems to increase the amount of posts.

Last, specific characters influence the distribution. Location of public transit station, breaking events, and hlgnet location, etc. could contribute to peaks in the above diagram as well. For instance, the Zhongguancun Life Science Park is the most popular employment base in category one. This can be explained by its large scale and long history. Nevertheless, the fact that one subway station on the Changping Line is named Zhongguancun Life Science Park also contributes to the daily experience category for many users to discuss about their commuter condition. Incidents like a traffic accident near Erbozi Industry Park may contribute to the small peak in frequency (35). And because the Huilongguan Community Forum's office locates in Beijing II Industry Park, many ads contain this location, which explains its relatively high frequency (21) with such a small area.

In the last two categories, the Shangdi IT Industry Base overwhelmingly exceeds others. Though the Shangdi IT Industry Base locates out of Huilongguan Neighbourhood, its high frequency parallel with the fact that majority of Huilongguan residents have a job relevant to information technology industry and most of such residents work in Shangdi.

54 Residents' attitude towards the development of employment bases

The following diagram shows the attitude to each industry parks.

120

Positiveattitude in Motion and proposals " Supporting narmtive of daily experience " Supporting reprint of news and articles

Negative attitude in motion and proposals 80 " Criticizing namtive of daily experience * Criticizing reprint of news and articles Frequency in informative messages Frequency in advertisements

40

Industry Base fndisiry Base -20

-40

-60

-80

Figure 18 Residents' attitude towards the development o employment bases

All first three categories have a positive to negative ratio that is larger than 1, which indicate that in general, netizens positive and supportive attitude exceeds the negative attitude.

When taking a close look at the negative posts and replies, complains mainly concentrate in topics of commuting. For example, Subway Changping Line formulate the station called Life Science Park, which links the main body of Changping New Town to the north, and connect the west part of Huilongguan to Subway Line 13 to the south. However, the location of Life Science Park Station isn't adjacent to the Zhongguancun Life Science Park. A 1.2 kilometres walk is needed from the station to the science park. This plan bothers many netizens both in daily commute and occasional use of subway line to the central city. Complains of bad experience in

55 daily commute like this reveal the fact that even these employment bases are developed, the inadequacy of coordinating and cooperating amenities could still leave problems.

In the discussion of whether an industry park has benefited residents in Huilongguan, 65% of people support the development because of the following reasons: o The development of such industry parks boost local economy and contribute to local fiscal revenue. This increment may help to facilitate urbanization process near Huilongguan and elevate local governments' administrative quality and reduce corruption. o The developments attract new population, which increase the demand of service industry. This will both benefit the existing commercial amenities, and will require amenity development which can also serve the current residents. o Good landscapes in industry parks like Zhongguancun Life Science Park become attraction for residents' daily exercise.

5.3 Top Discussions in the Online Forum

The top fifteen most popular discussions in opinion-oriented categories are summarized in the following table. Some topics directly discuss plans and developments of addition of employment bases; others don't produce arguments that are directly related to work zones, but produce relevant comment or express relevant opinions to these employment bases.

Table 7 Most popular topics regarding to employment bases

56 Topic Replies Views Category Ultimate resolution of Subway Line 13, 8, Changping and 16 67 2915 Motions and Proposals Call for comments: proposal for 2013 Political Consultative council 60 8602 Motions and Proposals The "Great Zhongguan" Plan: great possibility of the merging of Haidian and Changping 36 12319 Reprint of news articles Why not build more industry campus in Huilingguan? 34 2639 Motions and Proposals South area of Changping District will be included in "Great Zhongguan" Plan 15 818 Reprint of news articles Iluilongguan will absorb relocated population from central city's redevelopment 15 1699 Reprint of news articles TBD Plan will raise housing price in Huilongguan 13 5271 Reprint of news articles Ilaidian district control its permanent resident population within one million 13 5438 Reprint of news articles Exposure of negative behavior: someone jumping a queue on Line 8 this morning 12 3725 Narrative of daily experience Sany Industry move its headquarter to Huilongguan I1 562 Reprint of news articles Build lluilongguan-Tiantongyuan national community integrated demonstrate zone 10 1443 Motions and Proposals A proposal to optimize bus routes in Huilongguan 9 1133 Reprint of news articles Sany Industry's headquarter will locate in Huilongguan 9 2061 Reprint of news articles The Shangcheng community is determined as illegal project 9 797 Reprint of news articles Shangcheng project is one of them! Nine plots that use R&D landuse to build apartments 8 2311 Reprint of news articles are investigated and punished

The popular topics focus on the following aspects: o "Great Zhongguan" Plan: is an integration of administrative resources and industry park resources, to build a bigger platform for world-class high-tech industrial bases. The ratio of supportive opinions to oppositions about planning employment bases is 26/17. o Discussion of developing more industry campuses in Huilongguan. The ratio of supportive opinions to oppositions about planning employment bases is 16/4. o Proposals of optimizing traffic routes.

Counterviews on the work zone planning concentrate on questioning the plan's implementation, and potential burden the employment bases will spike on both environment and on public facility such as elementary school and hospital capacity.

The supporting netizens argue that their positive impact on local economy and their role in meeting demand of local job opportunity can beat the disadvantages.

5.4 Findings

General speaking, the development of employment bases in and near Huilongguan neighborhood relieve the demanding pressure of nearby work place in Huilongguan.

57 First, as time goes on, more and more scaled entrepreneurs choose to locate their headquarters and brands in the employment bases. Most of them are attracted by the government's favorable policies like tax abatement, and by the location advantage of high-tech agglomeration.

Second, these employment bases do absorb local workforce as well as introduce new employers into the neighborhood. Large amount of IT engineers choose to buy house units in Huilongguan for its convenient location and relatively low price. ((R23) I chose to leave here because the distancefrom here to my husband'workplace is really small.)((R]3) ... definitely most people live here because of the low housingprice.) At the same time, many young residents whose reason to live in Huilongguan is relocation or marriage find new skilled jobs in surrounded industrial parks. ((R15) after graduation, I tried to find ajob in Shangdi Industry park. It is convenient to live here and work in Shangdi). The employment bases' demands of workforce facilitate the demographic dynamic of Huilongguan in which the community gets fresh blood ((R26) I moved here to help my relative. Now I am responsiblefor the daily operation of his small grocery) ((netizen) Sany attractnuminous young pretty girls in the neighborhood). Meanwhile, the vigor of workforce also support more third industry such retail, dining service, recreational service, requirement of children' education and etc.

Third, the employment bases' influence vary base on preliminary planning decisions. Proximity and accessibility play an important role in it. Other elements such as built quality of environment, openness and publicity also influence the effect.

Fourth, the development of employment bases has improves the live work balance, yet not enough. Still, 65% of residents' work locates outside Huilongguan or above employment bases. Furthermore, the preliminary plan of Huilongguan left little space for work zone development, which badly restricts the increment of job opportunity within the neighborhood. Last, addition of employment bases needs parallel development of other amenities and facilities.

58 6 Commercial Amenities and Services

China, as a large economy, has been progressing rapidly in its urbanization. Among the hundreds of cities that have been growing rapidly, first tier cities like, Beijing, Shanghai, and Guangzhou have especially significant acceleration on their development speed. This occurs by taking advantage of their regional centre positions that allocate them with the largest amount of resources, including capital and manpower. The average income in these cities has been growing fast, therefore, people started to increase their expenditure in order to increase the quality of life.

When making decisions about where to live in Beijing, the only concern is not accessibility to employment. The amenity and infrastructure services became increasingly important criteria for the competitiveness of different residential community in Beijing. On the one hand, people want a better quality life, so they will look for a place that could provide them with enough amenities including food and beverage in addition to sport and leisure services. On the other hand, they are responsible for taking care of their families, which includes seniors and children, whether this area could have access to good hospitals or schools; these decisions weigh very much on the person's choice of home location.

Taking Huilongguan as a case study, to analyze and evaluate their amenity and infrastructure service is a very important step.

6.1 The Amenity Index

Thanks to information technology, open platforms currently allow us to able to access data. In this analysis we developed a method to evaluate Huilongguan's amenity and infrastructure service by measuring the index of different amenity or infrastructure component.

Data Platform.

The major data source used is the website called, Dazhongdianping (public review). It is an information platform for urban leisure services based on mass commentary. By the end of 2014, it already covered 300 cities in China, with service providers registered for over 4 million. This website has very convincing coverage in first tier cities,.

59 As mentioned in Chapter four, the Dazhongdianping search engine allows users to search by service type and area. The website returns a number of accessible services in different category for every query. For example, searching "Entertainment" in it, the website directs the user to a webpage showing the number of each "Entertianment" genre such as theaters, karaoke studios and board game bars that is accessible in Huilongguan.

Categories

Since Huilongguan is mostly functional as a residential area, the analysis will therefore be concentrated on four categories that are most closely related to residential amenity and infrastructure. The four categories are: food and beverage, leisure and entertainment, sports and health, education and hospital. Among them, the first three are categorized as amenity services, while the last is categorized as a public facility service.

The index

In order to understand how well Huilongguan is doing in providing sufficient amenity and infrastructural services, the entire city of Beijing, as a whole, is examined. Given a specific service type, taking food as an example, by dividing the overall number of food services (N=90103) by the total population (P=2 1,000,000) in Beijing, the average number of accessible food services per person in Beijing (n=4.29/1 000ppl) is calculated.

1 Nfood-Beijing/PBeijing= nfood-Beijing = 4.29/1000 pp

This number reflects the average condition in Beijing. Following that, by doing the same calculation, the same number for the Huilongguan area is calculated:

Nfood-HLG/PHLG =nfood-HLG = 3.55/1000 pp1

By dividing Huilongguan's number by Beijing's number, the Huilongguan Food service index I is calculated:

Ifood-HLG = nfood-HLG/nfood-Beijing = 0.83

60 In summary:

Table 8 Formula of the amenity index

Per Person Amenity Index Areal Accessible Amenities Beijing Accessible Amenities Per Person

A B B/A

Sejing Restaurant Uon Restaurant Index Wanglng Restaurant' Index I 10000 pp /1000 ppI 1000 ppI eawft 90103 4.29 1631 3.55 083 162 4.66 109 population 21000000 46600-6300

This number provides a sense of Huilongguan's progress in comparison to the city of Beijing. Regarding a specific type of service, if I >1, people living in Huilongguan have relatively higher accessibility towards this service, which tends to positively influence the overall performance of this area. If I < 1, Huilongguan have less accessibility to this service. In terms of food, it's obvious that people living here have less access to food services (restaurant, caf6, bakery etc.)

In the following analysis, this method will be applied to each of the four categories and their sub-categories to give the Huilongguan area a well-rounded evaluation and reading.

6.2 Comparative Evaluation - Wang Jing Area

In this analysis, given the fact that Huilongguan is a residentially oriented area, instead of only referencing it in comparison to the whole city of Beijing, it will be of more help to discover a comparative case.

Located in the Northeast part of Beijing, seated in between the 4 11 and the 5h Ring Roads, Wang Jing is one of the most well developed residential areas in Beijing, with a population around 350,000 people, including professions from all over the world, especially Korea.

In 2014, unit prices in the Wang Jing area reached 44,000 RMB/m2, which is significantly higher than Huilongguan's which reached 31,000 RMB/m2. Price differences were attributed by their locations. However, by comparing housing prices between Wang Jing and other residential areas that are at similar locations, it's obvious Wang Jing's housing price is still higher, which

61 -- ANN-

means people would like to pay a premium to stay in Wang Jing only to enjoy its amenity and facility services.

By putting together the amenity and facility indexes of Huilongguan and Wang Jing and observing their differences, some conclusions can be drawn regarding the developments of Huilongguan, in terms of what has been achieved and what is lacking.

WWV J Ing

OOR

40000 RMB

Figure 19 Three residential neighborhoods and their housing prices

6.3 Food and Beverage

Chinese people have developed a long history of food culture. By analysing the food service index as well as other data, a lot of information can be read:

Table 9 Index of different style of food

62 60 -,q 9estUwant/10000 p0 Hudowv"eR**tWvsenL wndas Waging Resaurant9 e -,g oo0 '56' 361 OA 0 19 0 568 0171 S-an F 6o189' 029 121 026 W 62 023 Sv, F ooe 2055 0 10 39 008 QW 34 010 No0,T E.a~. '846 009 5 0 13 23 007 C.&-tone Food '553 00? 12 003 34 010 6ve."n 004 24 00 om i6 005 Regional 1'S11e F 00 55? 0,03 5 00' 12 0031 ooFa" 530 003 a 002 1 000 Chinese X Food 436 0 02 $ 00 1 02 4 001 sna-30og Fcod 37 002 2 000 3 0011 ti~wFood 209 001 5 00 1,06 3 001 -vwFow '29 00' 5 001 4 001

Korer '32? 006 1 004 l 217 062 Regional "4 05 19 0 04 54 0151 Foreign sow!E,234 001 0 000, 7 0 02

Pas, F nod 3.54 46 06 1 V U 488 139 Sa1e1V 6266 030 95 0?' 141 040 "0 Px 6048 029 142 031 107 93 027 Other '00622 00 34 007 0.0 12 0031 Types S. Foo 125 006 32 007 1 16 23 007 SO9F 725 003 16 003 101 16 0 05 l ie 001 000

794 0 5 309 06 1s9 279 0 0 1M Toa 60*0 429 1631 355 (M 1632 466 0ODpA*s0I '000000 460000 350000

<0.5 0.5-1 1-1.2 1.2-2 0 >2

The Table above shows the number of all food categories in Beijing, Huilongguan, and Wangjing areas. The coloured columns represent the food index in Huilongguan and Wang Jing, where different levels of colour suggest the value of each index number.

Overall, the table tells that Huilongguan has a relatively low food index (0.83), and it's lower than Beijing average, whereas Wang Jing's index (1.09) is slightly higher than 1. The result suggests that Wang Jiang beat Huilongguan by over 30% for food service per person.

Looking more into the details in the indexes of subcategories, the categorization was interesting since it actually allowed for reading demographics of two places.

First of all, both of these two areas have very little Beijing cuisine (0.05), which highlights that both areas are composed mostly of immigrants. As for regional cuisine, Huilongguan is doing very well in the areas of North-eastern and North-western food. Wangjing has relatively more Cantonese food and Jiangzhe Food; Guizhou food doesn't count because the sample is too small. It has a significantly high index number in foreign cuisine, especially Korean food. As a matter of fact, foreign cuisine and Cantonese as well as Jiangzhe food are considered expensive and high end compared with other kinds of food.

63 More interesting data includes the bakery index in Huilongguan, which is extremely behind (0.69). It is beaten by Wang Jing (1.35) for almost 100 percent.

100% Buffet

90% Sea Food

Hot Pot 80% Other Types Bakery Fast Food 70% South Eastern Japanese 60% eKorean Western 50% IO Guizhou Food Hubei Food Shandong Food 40% *Xinjiang Food Yunan Food 30% FoJiangzhe Food

a North Western 20% Cantonese Food North Eastern

10% Hunan Food Sichuan Food Beijing Food 0% Beijing - HiLopggyanW- fgIng

Figure 20 Distribution of food in two neighborhoods

After examining the index of food services, it is also worth for us to look at price distribution of each give area:

Table 10 Price distribution in the two areas

Price Beijing Huilongguan Wangling 0-20 4984 21% 92 25% 68 10% 20-50 10025 41% 184 50% 261 40% 50-80 5597 23% 77 21% 203 31% 80-120 1930 8% 11 3% 77 12% 120-200 1035 4% 7 2% 31 5% >200 680 3% 0 0% 10 2% Total 24251 371 650 Average Price 57.5 39.3 61.0

64 60%

50%

40% -Beijing 30% Huilongguan

20% -- Wangjing

10%

0% 0-20 20-50 50-80 80-120 120-200 >200

This set of data represents the percentage of food services in each price range (RMB). In terms of overall distribution, the majority of food services are allocated in the 20-50 RMB price range Compared with the overall price distribution in Beijing, Huilongguan has a relatively larger percentage of mid and low cost restaurants (75% against 62%), while Wang Jing has more mid-high end restaurants. From the data, it shows that the curve doesn't really drop from the 20-50 to the 50-80 range. In terms of average price per restaurant, Huilongguan is significantly lower than the Wang Jing, which is about the same but slightly higher than the Beijing average.

Several conclusions could be drawn regarding Huilongguan from the food and beverage analysis:

1) As a residential area generally for immigrants, immigrants in Huilongguan are mostly from inland China, and possibly more from the northern part of China.

2)The current food and beverage market in this area is aimed at mid and low income persons, which could be detected from the price distribution.

3)Compared with more developed and high-end residential communities, Huilongguan needs more mid-to-high end restaurants, including foreign restaurants and bakeries.

Although this is a quantitative analysis, these conclusions are still slightly suggestive. Since food and beverage markets, in a specific area, are generated by the population and their income, in terms of development suggestion. Rather than investing to attract more mid-to-high end F&B service, it is more about attracting more high income population.

65 6.4 Leisure and Entertainment

Table 11 Leisure and entertainment index

Beijing Average/10000ppl Huilongguan Average/i index Wangjling Average/10Index Cafe 2691 1.28 29 0.63 ji WM 128 3.66,. Tea House 1853 0.88 27 0.59 0.67 41 1.17 1.33 massage 4505 2.15 119 2.59 1.21 142 4.06 bath 3190 1.52 121 2.63 30 M,86 05 KTV 1588 0.76 14 0.30 27 0.77 L2 Art&culture 3946 1.88 24 0.52 141 4.03 Cinema 315 0.15 5 0.11 0.72 a 0.23 Pool Game 1497 0.71 51 1.11 33 0.94 1.32 Table Game 235 0.11 7 0.15 1.36 4 0.11 1.0 Entertaining 817 0.39 14 0.30 0.78 22 0.63 Internet Bar 1255 0.60 17 0.37 0.62 19 0.54 0.91 Park 856 0.41 6 0.13 20 0.57 1.40

Total 31834 15.16 500 10.87 0.77 710 20.29 134

Population 21000000 460OW0 350000

In leisure and entertainment indexes, it can be easily observed that the gap between Huilongguan and Wangjing increased even more. Wang Jing's index is higher than Beijing in almost every aspect. The only two categories that Wang Jing is lower (bath and internet bar) actually reflect the fact that Wang Jing has more developed infrastructure, so that bath and internet services become privatized in comparison with Beijing.

Most of Huilongguan's indexes are lower than Beijing's average, which reflects the lag of leisure and entertainment development in this area. The higher index of the bath category proves that that the infrastructure in Huilongguan has yet to improve. The higher index of pool games and table games means that there are relatively more young people with middle or low income because these two consumption types are generally considered low cost.

66 8(V%

60%, Puo" Gajme

-Beijing -Huilonggun -Wangjlng

Figure 21 Leisure genre and distribution

One crucial thing that needs to be pointed out is the art & culture service index. Huilongguan, though labelled as a cultural residential area by the government, is actually an area that is seriously lacking art and cultural activities. Huilonguan's index in art and culture is significantly low (0.28), while at the same time Wang Jing's reaches an impressive 2.34, there is almost 10 times difference.

As a matter of fact, in order to develop into a better residential town for Beijing, the art and culture sectors are definitely worth investing in to cater to future development. Here, unlike food and beverage, which is more driven by existing populations, the art and culture market could actually attracting people to come and settle. Therefore, in terms of developmental strategy, it's strongly suggested that Huilongguan invest to attract more art and culture services, and implement them into the community.

6.5 Sports & Health

Unlike the previous two indexes, the overall amenity index of sports and health in Huilongguan (1.15) is slightly higher than the average of the entire Beijing area. However, when comparing the indexes between Huilongguan and Wangjing, the dramatic difference indicates that further development is supposed to be implemented in terms of sports and health.

67 Table 12 Sports index comparison

Hulene wanftg keng population per 10000 population per 10000 population per 10000 460000 3000D0 21000000 SPORS & HATH Snooker & Pool 48 1.04 33 1.10 1497 0.71 Gym 44 0.96 49 1.63 1937 0.92 15 0.33 23 0.77 641 0.31 Bedminton 12 026 10 0.33M 291 0.14 Dance 12 0.265 38 1.27 691 0.33 martial arts 9 0,20 6 0,20 360 017 Tennis 8 0.17 Lit 6 0.20 1 315 0.15 Swimming 7 0.153 22 0.73 535 0.25 7 015 10 0.33 213 0.10 Baskeba . Footbal 5 0.M1 S 0.17 132 0.06

Total 167 3.630435 S 202 6.733333 6612 3.148571

By examining more details at the index of subcategories, it's clear that badminton and football are the only two index values higher than 1.50. Obviously, the Huilongguan community has a tradition of annual inter-league matches of these two sports. Thus, these two sports are relatively more popular among Huilongguan residents. It can be concluded that the internal demand does play a crucial role in the development of sports and health amenities.

R~1

"'V

nlol wts' 4~I

4Th~ MDv

01, A~ a( 6P'o

*S tAe & i'ovt

Beijing Hudongguan Wanglpig

Figure 22 Sports genre and distribution

The proportion graph shows that the amenity distribution of the Huilongguan area shares similarities with the entire Beijing area. Compared with Wang Jing, the amenity distribution in Huilongguan is more generic without any specific preference. Therefore, in order to enhance the diversity and the density of sports and health amenities, Huilongguan is encouraged to organize more, various public events and advocate residents to participate in different types of sports.

68 6.6 Public Services

China's one child policy causes all parents to care and to be concerned very deeply about their child's development. Therefore, the quality of schools has significant impact on parents' choice of community. In some extreme cases, parents are even willing to sacrifice the convenience of working from home in order to live closer to a good school.

Table 13 Infrastructure and public service index

Huilongguan Wanjing Beijing population per 10000 population per 10000 population per 10000 460000 300000 21000000 EDUCATION kindergarten 101 2.20 58 1.93 3924 1.87 primary school 15 0.33 18 0.60 0.93 1523 0.73 middle school & high school 4 0.09 9 0.30 0.84 753 0.36

PUBUC SERVICE Hospital 99 2.15 0.3 75 2.50 Mas 6203 2.95 Bank 81 1.76 229 7.63 11954 5.69

Huilongguan's infrastructure can be defined as the most underdeveloped amenity sector out of the entire family. In terms of education, there is a dramatic gap between preschool education (1.14) and primary-school education (0.45). The graph shows that the kindergarten index in Huilongguan appears to be in a good condition, while the school index is surprisingly low compared with Wang Jing or the Beijing area. The situation becomes increasingly worse when the education level gets increases. The Huilongguan neighbourhood is not able to provide sufficient educational resources. In this case, many families start to migrate to other places when the existing education environment cannot fulfil their basic requirement.

To transform the Huilongguan area into a sustainably living neighbourhood, the government should take more actions to increase supply and improve quality of the nine-year compulsory education.

6.7 Findings

Ten years ago, while Huilongguan's amenity plan hadn't been fully implemented, the then existing amenity space only had a occupancy rate of 50% (Chenglei Li, 2006). News articles reported that the early inhabitants suffered from scarcity of all kinds of amenities. "This kind of plan has only residential buildings with no industry and no infrastructure. Certainly a lot of

69 problems came up. The ideal new town has facility developed with housing. However, facilities formation and accumulation requires patience, time and cost. (Tang, 2007)"

Figure 23 photos of amenities in Huilongguan, in 2009

Comparing to the current condition of amenities in Huilongguan, improvements have emerged. The neighborhood has a more mature system of commercial amenity. Gaps between Huilongguan's amenity level and Beijing's decrease or even are overtaken. Interviews echo this finding:

"(R5) the commerce is complete and convenient, and all commercial amenities distributed adjacent to home. We usually went to the Hualian mall. It is really convenient, with a cinema and lots of restaurants....speak of other recreationalamenity, eh; you must go to the city center."

"(R25) yes, the commerce is adequatefor daily life. And there are many outdoor athletic facilities in housing clusters. We also take walks to HuilongguanSports Parkfor exercise."

"(R14) restaurantsand recreationalamenities are adequate. They are much improved then before. However, it is so inconvenient that there are no banks nearby. And if have serious illness, I will choose hospitals in town, because I don't quite trust the medical skill of hospitals in Huilongguan."

The research indicate that service level of commercial amenity and facility has been improved to satisfy residents' basic requirement of daily life. Among all categories, however, the most important two elements are cultural amenity and educational facility. The former will significantly increase the quality of life in this area and the latter can convince more families to stay in this are in the long run. Despite the lower housing price, Huilongguan's core

70 competitiveness comparing with other residential neighbourhoods like Wang Jing weakens because of the scarcity of high-end commercial amenities like cultural venues and advanced public facilities like secondary schools and comprehensive hospitals.

71 7 Conclusion

7.1 Research Results

Let us re-examine the original research question. Can crowd sourced data be used to evaluate planning strategy and facility quality? After analyzing this method, this thesis provides a positive answer to this question. By collecting data from a wide range of online community forums and social media websites, as well as analyzing keywords of distribution and sentiment, the planning of public facilities and amenities in Huilongguan has been investigated and evaluated. This can be one of many tools to evaluate the planning of amenities.

In order to sharpen the analysis process into an operational tool, the research steps of previous chapters are summarized below.

1. Data Resources

a) Choosing resources - in an era that features the rapid development of Internet technology and social media, as well as the popularization of mobile terminals, online communication has become cheap and widely accessible. Furthermore, online community forums have sprung up in different regions like mushrooms. In China, many of them, such as the Huilongguan Community Forum, have even developed into first-class online forums. Because they are regional (or subdivided geographically like Dazhongdianping) and feature widespread participation by local residents, social data extracted from these online platforms is a major resource.

b) Grabbing data - technology such as web crawlers (also called web spiders or automatic indexers; scripts that systematically browse the Internet and grab data for later processing) area used in the collection of data. Queries are made in the search interface using the names of research objects or keywords.

c) Structure of results - information such as discussion topics, time, popularity

72 (amount of views) and replies can be collected for later analysis.

2. Data Processing

Data is processed to merge repeat records and remove irrelevant ones.

3. Content Analysis

a) Classification of data - this step helps to explain which topics related to amenities have drawn attention. In Chapter 5, a number of articles about new plans or proposals, as well as suggestions, are discussed. Narratives about daily experiences are recorded to reveal the level of satisfaction and ads are posted to show the influences of amenities on daily life. Similar to Chapter 5, data is categorized into three parts in Chapter 6. These categories are discussions related to plans, built facilities (subway system hardware) and daily operations and services (software).

b) Distribution of topics - this step reflects the daily influence of amenities on the lives of residents.

c) Attitude and satisfaction - this step indicates whether the general attitude towards an amenity is positive or negative. Major concerns and the reasons that have led to discontent are analyzed in detail.

d) Case study of the top debates on online forums - the in-depth study in this part analyzes the topics with the most participants and reveals embedded facts.

4. General Evaluation and Conclusion

A general evaluation of amenities is conducted by synthesizing the above steps.

In summary, a duplicable method is proposed for planners and decision makers to gain a practical understand of the planning process for amenities.

7.2 Major findings

73 How can this tool be used? What are the findings from using this method? Echoes the second research question, findings can be extracted from the above-mentioned process of analysis and additional amenities can be evaluated. This evaluation reflects the partial improvement, as well as reveals the problems, of amenities. a) In general, local residents hold a positive attitude towards the improvement of employment bases, transportation and commercial amenities. b) All three types of amenities were improved, but these improvements were insufficient. They satisfied the basic requirements of residents. However, the developments of advanced facility and amenity such as high quality secondary schools and comprehensive hospitals are inadequate. c) Although the overall plan benefited the neighborhood, the lack of detailed design caused major dissatisfaction. For example, the lack of noise-reducing walls for the subway Line 13 annoyed residents who live along this subway line. Additionally, the lack of screen doors in subway station platforms was a safety hazard, and even resulted in accidents involving passengers falling on to the subway tracks. d) Furthermore, after the plans for amenities were implemented, daily operations and maintenance became big issues that were subject to complaints and criticism. e) The construction of one type of amenity required parallel and matching construction of others. For example, a new industry park required a corresponding transportation route and the new subway station required small amenities such as deli vendors to accommodate passengers.

Employment bases

To some extent, the construction of employment amenities provided new employment opportunities and new locations for residents of neighborhoods in Huilongguan. However, the preliminary plan did not leave sufficient land for the development of commercial or office amenities, which restricted the development of a multi-functional district.

74 Mass Transit

Over the past decade, a large amount of new transportation infrastructure was built. Problems with traffic congestion were solved and commute times were reduced. However, improved traffic only cured the symptoms, not the disease. The key was planning a multi-functional neighborhood during preliminary stages, or at least leaving sufficient land for further commercial and industrial development.

Commercial Amenity and other facility

Despite the fact that current commercial amenities meet the daily requirements of dining, shopping and recreation, the scarcity of scaled and quality amenities weakened the core competitiveness of the neighborhood. It caused residents to leave the city center for better education opportunities for their children.

The existence of Huilongguan, as a suburban mega community in Beijing, came from necessity. Over the past 15 years, Huilongguan contributed heavily to absorbing Beijing's urbanization, as well as providing a strong base for the city to develop its IT industry. With Beijing's expansion towards the northwest, Huilongguan is a welcoming gentrification opportunity to be upgraded to a more attractive neighbourhood. However, in order to seize this opportunity, the government must provide more high-income jobs that are accessible to this area as well as invest in the amenity and facility development.

7.3 Limitation and Future Research

First, the social data extracted from online forum are biased. In crowdsourcing, a generic element is who constitute the active crowd. This question isn't discussed in the thesis. Though data are completely analyzed (e.g. employment bases and commercial amenities) or randomly sampled (mass transit), the residents who participate in online forum may not represent the whole group of people in the neighborhood. If an analogy is drawn between the abovementioned data collecting process and the sampling process in statistics, there may be biases. Main body of active netizens in hlgnet is comprised of early inhabiting, high-income, and well-educated people. On the contrary, opinions from

75 renters (mainly low-income labors in service industry) are barely expressed from the online forums and public review networks. A seventh of the interviewees use the Huilongguan Community Forum, while the ratio reaches 6/14.

Second, the research collects prolonged data from the very beginning of the online platforms. This research builds its comparison between social data and historical reality extracted from literature. Neither scopes of time are specified, nor are before-and-after situations compared from same data source. Despite cost, longitudinal studies such as penal study can be conducted in the future tract the dynamic of Huilongguan's amenity.

Third, this method is only used in Huilongguan's case to conduct planning evaluation. However, plan concept and social context may distinguish, and the applicability of this method need to be testified in future research.

7.4 Implications

Social data from community forum and public reviews is one tool among numerous others to evaluate planning intervention. Moreover, this tool provides a more transparent and less time-consuming way to analyze first-hand data in order to assess the outcome of planning. New communities such as Huilongguan are prevailing in not only mega cities but also in many second-tier cities in China. Moreover, many of these new communities face similar problems as Huilongguan, including insufficient local amenities and facilities, excessive commute times, traffic congestion, etc. By evaluating the development of amenity in Huilongguan, effective methods can be discovered for decision makers and planners to alleviate such problems.

76 8 Bibliography

Baer, William C. "General Plan Evaluation Criteria: An Approach to Making Better Plans." Journal of the American PlanningAssociation 63, no. 3 (September 30, 1997): 329-44.

Bentley, John P., Cynthia J. Brown, Gerald McGwin, Patricia Sawyer, Richard M. Allman, and David L. Roth. "Functional Status, Life-Space Mobility, and Quality of Life: A Longitudinal Mediation Analysis." Quality ofLife Research: An InternationalJournal of Quality of Life Aspects of Treatment, Care and Rehabilitation 22, no. 7 (September 2013): 1621-32.

Bunnell, Gene, and Edward J. Jepson. "The Effect of Mandated Planning on Plan Quality." Journal of the American PlanningAssociation 77, no. 4 (October 1, 2011): 338-53.

Conroy, Maria Manta, and Philip R. Berke. "What Makes a Good Sustainable Development Plan? An Analysis of Factors That Influence Principles of Sustainable Development." Environment and PlanningA 36, no. 8 (2004): 1381-96.

Jun, Myung-Jin, and Jae-Wan Hur. "Commuting Costs of 'leap-Frog' Newtown Development in Seoul." Cities 18, no. 3 (June 2001): 151-58.

Lee, Sooil, Seungjae Lee, Hyeokjun Son, and Yongjin Joo. "A New Approach for the Evaluation of the Walking Environment." InternationalJournal of Sustainable Transportation7, no. 3 (May 2013): 238-60.

Quality-of-Life Community Indicatorsfor Parks, Recreationand Tourism Management. Accessed December 14, 2014. http://www.springer.com/social+sciences/wellbeing+%26+quality-of-life/book/978- 90-481-9860-3.

Takemi Sugiyama, Jacinta Francis. "Associations between Recreational Walking and

77 Attractiveness, Size, and Proximity of Neighborhood Open Spaces." American JournalofPublic Health 100, no. 9 (2010): 1752-57.

Tan, T. "A Mathematical Model for Commuter Traffic in Satellite Towns." TransportationScience 1, no. 1 (February 1, 1967): 6-23. doi: 10.1 287/trsc. 1.1.6.

Gibson, Neil. "The Privatization of Urban Housing in China and Its Contribution to Financial System Development." Journalof Contemporary China 18, no. 58 (January 1, 2009): 175-84.

Li, Xu Jia. "A Study of Beijing's Suburban Residential Areas." Advanced Materials Research 250-253 (May 2011): 2418-23.

Scott, Andrew. "Design Principles for Sustainable Urban Housing in China." In Sustainable Urban Housing in China, edited by Leon Glicksman and Juintow Lin, 24-43. 2006.

Shaw, Victor N. "Urban Housing Reform in China." HabitatInternational 21, no. 2 (June 1997): 199-212.

Suburban Beiing - University of Minnesota Press. Accessed April 29, 2015. https://www.upress.umn.edu/book-division/books/suburban-beijing.

Wang, Ya Ping, and Alan Murie. "Social and Spatial Implications of Housing Reform in China." InternationalJournal of Urban and Regional Research 24, no. 2 (June 1, 2000): 397-417.

. "The Process of Commercialisation of Urban Housing in China." Urban Studies 33, no. 6 (June 1, 1996): 971-89.

Wu, Fulong. "Changes in the Structure of Public Housing Provision in Urban China." Urban Studies 33, no. 9 (November 1, 1996): 1601-27.

Yang, Jun, and Zhou Jinxing. "The Failure and Success of Greenbelt Program in Beijing."

78 Urban Forestry & Urban Greening 6, no. 4 (November 15, 2007): 287-96.

Zhou, Min, and John R. Logan. "Market Transition and the Commodification of Housing in Urban China*." InternationalJournal of Urban and Regional Research 20, no. 3 (September 1, 1996): 400-421.

Ellison, Nicole B., Charles Steinfield, and Cliff Lampe. "The Benefits of Facebook 'Friends:' Social Capital and College Students' Use of Online Social Network Sites." Journalof Computer-MediatedCommunication 12, no. 4 (July 2007): 1143-68.

Nahapiet, Janine, and Sumantra Ghoshal. "Social Capital, Intellectual Capital, and the Organizational Advantage." Academy ofManagement Review 23, no. 2 (April 1998): 242-66.

Portes, Alejandro. "Social Capital: Its Origins and Applications in Modem Sociology." Annual Review of Sociology, 1998.

Valenzuela, Sebastian, Namsu Park, and Kerk F. Kee. "Is There Social Capital in a Social Network Site?: Facebook Use and College Students' Life Satisfaction, Trust, and Participation." Journalof Computer-MediatedCommunication 14, no. 4 (July 2009): 875-901.

Den Hartog, Harry. Shanghai New Towns: Searchingfor Community and Identity in a Sprawling Metropolis. 010 Publishers, 2010.

Ducom, Estelle. "TAMA NEW TOWN, WEST OF TOKYO: ANALYSIS OF A SHRINKING SUBURB." Accessed November 2, 2014. https://halshs.archives-ouvertes.fr/halshs-00203107/.

Eom, Sangkeun, and Seungil Lee. "Application of the Urban Green Assessment Model for the Korean Newtowns." In ComputationalScience and Its Applications - ICCSA 2009, 148-62.

Guo, Ming. "From Village to Small Town in Contemporary China: The Transformation

79 of Civic Space." Thesis, Massachusetts Institute of Technology, 2005. http://dspace.mit.edu/handle/1 721.1/34175.

HUANG, Sheng-li, and Yue-ming NING. "The New Town Movement And Its Implications." Urban Research 2003, no. 4 (April 2003). http://en.cnki.com.cn/Articleen/CJFDTOTAL-XDCS200304002.htm.

Jun, Myung-Jin, and Jae-Wan Hur. "Commuting Costs of 'leap-Frog' Newtown Development in Seoul." Cities 18, no. 3 (June 2001): 151-58.

Kiuchi, Shinzo, and Noboru Inouchi. "New Towns in Japan." Geoforum 7, no. 1 (1976): 1-12.

Merlin, Pierre. "The New Town Movement in Europe." The ANNALS ofthe American Academy of Politicaland Social Science 451, no. 1 (September 1, 1980): 76-85.

Qu, Yan, Chen Huang, Pengyi Zhang, and Jun Zhang. "Microblogging After a Major Disaster in China: A Case Study of the 2010 Yushu Earthquake." In Proceedingsof the ACM 2011 Conference on Computer Supported Cooperative Work, 25-34. 2011.

Dunne, Kerill. "Can Online Forums Be Designed to Empower Local Communities?" TripleC (Cognition, Communication, Co-Operation): Open Access Journal for a Global Sustainable Information Society 9, no. 2 (July 2011): 154-74.

Farisi, Mohammad Imami, imamfarisigut.ac.idimamfarisigyahoo.com. "Social Relation Networks in Ut-Online Community Forum." Turkish Online Journalof Distance Education (TOJDE) 13, no. 2 (April 2012): 48-57.

Gerlock, Jennifer Ann 1, and Dawn Lorrainel, [email protected] McBride. "Managing Online Discussion Forums: Building Community by Avoiding the Drama Triangle." College Teaching 61, no. 1 (Winter 2013): 23-29.

Gill, Preetinder Singh, and Billy Whisnant. "A Qualitative Assessment of an Online Support Community for Ovarian Cancer Patients." Patient Related Outcome

80 Measures 3 (December 2012): 51-58.

Guoxin Lii, Xue YangI, and Songshan (Sam) Huang2. "Effects of Social Capital and Community Support on Online Community Members' Intention to Create User-Generated Content." Journal of Electronic Commerce Research 15, no. 3 (August 2014): 190-99.

Laing, Angus, Debbie Keeling, and Terry Newholm. "Virtual Communities Come of Age: Parallel Service, Value, and Propositions Offered in Communal Online Space." Journalof Marketing Management 27, no. 3/4 (March 2011): 291-315.

Li, Limei, and Si-ming Li. "Becoming Homeowners: The Emergence and Use of Online Neighborhood Forums in Transitional Urban China." HabitatInternational 38 (April 1, 2013): 232-39.

Margerison, JillI. "Online Discussion Forums in the Classroom: Can the Principles of Social Media Benefit Literacy and Enhance Engagement with Learning?" Literacy Learning: The Middle Years 21, no. 2 (June 2013): 19-28.

Petri', Gregor, and Andra Petrov'i'. "Individual and Collective Empowerment in Online Communities: The Mediating Role of Communicative Interaction in Web Forums." Information Society 30, no. 3 (June 5, 2014): 184-99.

Rosenberry, Jack. "Virtual Community Support for Offline Communities Through Online Newspaper Message Forums." Journalism & Mass Communication Quarterly 87, no. 1 (Spring 2010): 154-69.

Tabor, Aidan S., and Taciano L. Milfont. "We Are All in the Same Boat: How Online Communities Facilitate the Process of Migration." New ZealandJournal of Psychology 42, no. 1 (March 2013): 31-35.

Zamani, Navid J., Graham S. Smith, and Gerald Monk. "Online Forums as Definitional Ceremonies." Journalof Systemic Therapies 32, no. 4 (Winter 2013): 1-18.

81 Al-Maimani, MAQBOOL, NAOMIE2, and AHMED M. "Semantic and Fuzzy Aspects of Opinion Mining." Journal of Theoretical& Applied Information Technology 63, no. 2 (May 20, 2014): 330-42.

Gunter, Barrie, Nelya Koteyko, and Dimitrinka Atanasova. "A Market-Relevant and Reliable Measure of Public Feeling?" InternationalJournal of Market Research 56, no. 2 (March 2014): 231-47.

Kaur, Jasleen and Jatinderkumar. "An Analysis of Opinion Mining Research Works Based on Language, Writing Style and Feature Selection Parameters." International JournalofAdvanced Networking & Applications, November 15, 2014, 7-13.

Lingyu Wang, Weina Qu, Xianghong Sun, and P.l.p. Rau. "An Analysis of Microblogging Behavior on Sina Weibo: Personality, Network Size and Demographics." edited by P.i.p. Rau, Vol. pt.I. Springer Verlag, 2013.

Liu, Bing. "Sentiment Analysis and Opinion Mining." Synthesis Lectures on Human Language Technologies 5, no. 1 (May 23, 2012): 1-167.

Pang, Bo, and Lillian Lee. "A Sentimental Education: Sentiment Analysis Using Subjectivity Summarization Based on Minimum Cuts." In Proceedingsof the 42Nd Annual Meeting on Associationfor ComputationalLinguistics. 2004.

Pang, Bo, Lillian Lee, and Shivakumar Vaithyanathan. "Thumbs Up?: Sentiment Classification Using Machine Learning Techniques." In Proceedingsof the ACL-02 Conference on Empirical Methods in NaturalLanguage Processing- Volume 10, 79-86, 2002.

Qu, Yan, Chen Huang, Pengyi Zhang, and Jun Zhang. "Microblogging After a Major Disaster in China: A Case Study of the 2010 Yushu Earthquake." In Proceedingsof the ACM 2011 Conference on Computer Supported Cooperative Work, 25-34. 2011.

Xuemei Zhou, and Yizhe Liu. "Data Detection, Tracking and Sentiment Analysis Based on Micro-Blog Data." Advanced MaterialsResearch 1079-1080 (November 11,

82 2014): 609-13.

Yoosin Kimi, and Seung Ryul Jeong. "Opinion-Mining Methodology for Social Media Analytics." KSII Transactionson Internet & Information Systems 9, no. 1 (January 2015): 391-406.

Ik'j, and ' "Em' 1-L-T1TA}%iMJ AX9 W J # 93r. M A tlU, no. 03 (2008): 56-61.

, and Lt. "5KMrY iT)WI t-* A EfL i) xM

1R." WiT E, no. 06 (2013): 117-22.

M. "AL ANiN At " i, 4 LALc,, 2010.

OL+*. "R*Aii 1A A? A R4t1cik 1JWiXhah9- t-N T_ I, RE f f, 4iA . ?,Ii, no. 08 (2004): 20-25.

EA. ."1L ib tk& el!iu]iJ R." Ek , A' , 2006.

49e, NM Ala and $$t.E~ JMMN~tKREi] $4 N81J." AkZItdJ, no. 03(2013): 27-33+113.

-EM. "4L b2$) etg ai-it ." f , di 2012.

# , and -'. k+M, MH h4qIj XL L l-E EflIR M It-EP It."A , no. 05 (2013): 24-30.

N. "$W~tditi6%K@K@R4%ii" ~, R**,2007.

10 (2009): 300-302.

jLX . " iN i-& I " 1 N YEtx t 0, no. 06

83 (2011): 41-47.

-h A . "ff a F m - 3ift A ) - A-Ikt$}k An It Rht rp. NR 3, tl M, , T 2005.

, 4 %2?,and ,). " IL Z F)QI $ A AW 4 Ek1W." - H !l)ifu-k*IK($1$84*fd), no. 04 (2009): 88-92.

4 , and WHAf*." A L Tfi i Y ) (M I M K I I1 T f@1 IT iTi fpl ," no. 07 (2007): 55-59.

84 Appendix I

The most popular posts regarding to employment bases

1. There is a great opportunity that Haidian and Changping will be integrated into one district, 'Zhongguan District'

a) Summary of News 2009/12/29

Soon after, the Zhongguancun National Innovation Demonstration Zone will appear as a special "company" integrated industry parks and employment bases across administrative region 'Haidian' and 'Changping' . Next year, the Beijing City will focus on the pooling of resources of Haidian and Changping Districts, integrating regional resources such as the core area of Zhongguancun and the Changping Industry Town, and the develop will focus on the construction of the northern part of the high-tech industrial belt which comprise the future increase of low-carbon, biological medicine, new energy, new materials, electronic information and other high-tech strategic industries.

b) Replies Summary: Views: 12330; Replies: 51; -/+: 20/12

Most popular opinion and amount:

o Negative: Questioning on authenticity of the news and questioning on implementation. 13; Positive: it is a process/ good beginning. 6. o Negative: Spiking Burden of children's education (elementary school). 5; Positive: potential opportunity for introducing new quality schools and reduce commuter distance to elementary schools. 3. o Negative: Stimulating housing price in Huilongguan. 2. Positive: 0. o Burden existing traffic: 2. Positive: 0. o Negative: 0; Positive: facilitating urbanization and governments' administration quality/ reduce corruption. 3.

2. 'Great Zhongguan' Plan will raise the housing price in Huilongguan

85 a) Summary of News 2011/11/18

As the first year , twelve five, , in 2011, new policies, new directions, all will be gradually clear with the successful open of the National People's Congress, among which is the "Great zhongguan" plan that has draw various attention for a long time.

According to the municipal plan, in the next five years, Changping District will construct the core area of Zhongguancun National Innovation Demonstration Zone, Zhongguancun Life Science Institute, Zhongguancun National Engineering and Technology Innovation Base, as well as coordinate with Haidian District to promote the building of Technology Business District (TBD), the Future Science and Technology Town, Shahe University City and other significant industrial function area. Finally in Beijing, the plan aims to create a world-class high-tech industrial base in north Beijing. The plan expects to fully realize the great idea of "the Great Zhongguan" which build on the platform of two districts, three towns and houndreds of schools.

This vision not only will integrate traffic, amenity and industry resource from Haidian and Changping and make them mutually reinforcing, but also will produce a broad and profound impact on whole city's economy, culture, and the development of high-tech industry.

b) Replies Summary: Views: 8306; Replies: 14; /-: 6/5

Most popular opinion and amount:

Positive: facilitating the economic development of Huilongguan/ increasing housing price in Huilongguan. 5; Negative: it is propaganda./ housing price only relate to demand-supply equilibrium./ The trend of industry is saturated. 6.

3. Why not build more industry campus in Huilongguan?

a) Summary of post 2010/07/27

What if Huilongguan develop some office buildings and technology campus? They

86 people can choose their job location close to their home, which will ease the traffic pressure. And everyone will be happy with the convenient employment location.

With the special loacation and human resources, Changping District Government should consider develop new science and technology parks, instead of building such parks in remote location such as Zhongguancun Changping Science Park which developed extremely slowly.

Someone may debate that this will introduce new population that ever more burden mass transit and traffic on roads. However, with these adjacent employment base, commuting demand will decrease. Moreover, with the influence of big enterprises, local government will pay more attention to impprove the traffic condition.

b) Replies Summary: Views: 2644; Replies: 34; +1-: 8/2

Most popular opinion and amount:

Positive: reduce the separation of work and residence./ effectively use public transportation and amenity resources, boost local business. 8; Negative:make no sense.

4. Build Huilongguan-Tiantongyuan national community integrated demonstrate zone

a) Summary of post 2010/07/27

What if Huilongguan develop some office buildings and technology campus? They people can choose their job location close to their home, which will ease the traffic pressure. And everyone will be happy with the convenient employment location.

With the special loacation and human resources, Changping District Government should consider develop new science and technology parks, instead of building such parks in remote location such as Zhongguancun Changping Science Park which developed extremely slowly.

87 Someone may debate that this will introduce new population that ever more burden mass transit and traffic on roads. However, with these adjacent employment base, commuting demand will decrease. Moreover, with the influence of big enterprises, local government will pay more attention to impprove the traffic condition.

b) Replies Summary: Views: 2644; Replies: 34; /-: 8/2

Most popular opinion and amount:

Positive: reduce the separation of work and residence./ effectively use public transportation and amenity resources, boost local business. 8; Negative:make no sense.

88 Appendix II

The most popular posts regarding to mass transit

1. Proposal of optimization of subway line.

a) Summary of News 2010/07/26

Ultimate optimization of mass transit: Line 13, Line 8, Line Changping and Line 16.

To ease the traffic congestion and to make residents life more convenient, the post bring forward the following proposal. o Line Changping: Add a new stop near existing Huilongguan Station and add a transfer station with Line 13 in Longze Station. o Line 13: do not bifurcate the Line 13. o Line 8: overlap its route with Line Changping within the 'Life Science Park' to 'Xierqi' section. o Line 16: no possible adjustment to fulfill Huilongguan residents' demand.

b) Replies Summary: Views: 2930; Replies: 67; -/+: 4/17

Most popular opinion and amount: o Positive: strongly supportive. / believing in the proposal's acception by Beijing Municipal Commission of Urban Planning and proposal's emplimentation. Negative: 4. o Some other opinions of the mass transit route.

2. Beijing Changping Line ' Xi'erqi' subway station will be put into use.

a) Summary of News. 2010/07/09

89 The Subway Line Changping Line successfully opened its trial operation this morning (2010/07/09).

Changping line starts north from Tombs scenic, extends south to the Light Rail Line 13 with a transfer station of Xi'erqi Station.

Changping line was constructed in two phases. The first phase links the South Station and Xi'erqi Station, whose overall length is 21.094 km and contains 7 stations.

Phase I splits the existing Line 13 in Xi'erqi Station into two sections, of which the western part of Xi'erqi-Xizhimen section of Changping Line is under thorough operation, while the eastern section of Xi'erqi Station to Dongzhimen station will form a new line 13.

b) Replies Summary: Views: 3496; Replies: 46; -/+: 19/2.

Most popular opinion and amount:

Negative: the Line 13 will be more crowed. / increase unnecessary transfer. /Strongly disagree the split of Line 13. 19. Positive: 2.

3. Enhancing the development of transit bus routes and commuter bus routes which link the major subway station. 2015/01/19

a) Summary of News.

Many locales are in a situation of broad roads with no bus routes which make it so inconvenience for residents to commute. Thereby I propose add one subway station in Line 8 between Xi'erqi and Longze, which will ease the pressure of Xi'erqi station.

And I propose to open new bus route that link local communities to nearby subway station, which will facilitate mode share of public transport and low-carbon trips.

Replies Summary: Views: 1042; Replies: 21; -/+: 1/15

90