Advances in Engineering Research (AER), volume 142 International Conference for Phoenixes on Emerging Current Trends in Engineering and Management (PECTEAM 2018)

A Survey: Data Mining Techniques for Social Media Analysis

1Elangovan D, 2Dr.Subedha V, 3Sathishkumar R, 4 Ambeth kumar V D 1Research scholar, Department of Computer Science Engineering, Sathyabama University, India 2Professor, Department of Computer Science Engineering, Panimalar Institute of Technology, Chennai, India 3Assistant Professor, 4Associate Professor, Department of Computer Science Engineering, Panimalar Engineering College, Chennai, India [email protected],[email protected], [email protected],[email protected]

 Abstract—Data mining is the extraction of present information databases. The overall objective of the data mining technique from high volume of data sets, it’s a modern technology. The is to extract information from a huge data set and transform it main intention of the mining is to extract the information from a into a comprehensible structure for more use. The different large no of data set and convert it into a reasonable structure for data Mining techniques are further use. The social media websites like Facebook, twitter, instagram enclosed the billions of unrefined raw data. The I. Characterization. various techniques in data mining process after analyzing the II. Classification. raw data, new information can be obtained. Since this data is III. Regression. active and unstructured, conventional data mining techniques IV. Association. may not be suitable. This survey paper mainly focuses on various V. Clustering. data mining techniques used and challenges that arise while using VI. Change Detection. it. The survey of various work done in the field of social network analysis mainly focuses on future trends in research. VII. Deviation Detection. VIII. Link Analysis. Index Terms— Data Mining, Social media, Social network IX. Sequential Pattern Mining. Analysis, Web Mining. Social network finds its application in several business activities like Co-innovation, Customer service, General I. INTRODUCTION promoting, increasing spoken promoting, marketing research, ata mining is a powerful tool which will facilitate to seek plan generation and new development, publicity, worker D out hidden patterns and various relationship between the communication and reputation management. data. Data processing discovers hidden facts from massive II. LITERATURE SURVEY

ALGORITHM S.NO AUTHOR PAPERS TITLE ADVANTAGES/LIMITATIONS / TOOLS . For various group of users, dissimilar interaction patterns that are not Identifying User similar can be experiential. Clustering 1 Marcelo Maia, et.al Behavior In Online . Characterize and identify the user Algorithm Social Networks profiles in online social networks. . Display more suitable advertisements based on user behavior  Using maximum group and sub grouping criteria social network Mohamad Al- Analysis Of Social structure are clustered. Cleavage Ant 2 Fayoumi, Souyma Network Using Cleaver Colony Metaphor  Imposing performance on test run Banerjee, et.al Ant Colony Metaphor using the standard group sequence data particularly for the clustering of the instance named as Keller 6.

Copyright © 2018, the Authors. Published by Atlantis Press. 109 This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). Advances in Engineering Research (AER), volume 142

 To perform analysis on public data Massive Social from a micro blogging network like David Ediger Karl Network Analysis: Graph Ct Snag 3 twitter. Jiang, et.al Mining Twitter For Algorithm Social Good  focus of analysts more on a much smaller data subset. Identification Of User Pattern In Social Decision Tree 4 A.SemanBozkur et.al  most precise results Networks By Algorithm Datamining Techniques SENTIWORDNET  relative improvement of 19.48% for 3.0: An Enhanced StefaoBaccianella SENTIWORD- the ranking by positivity and a 5 Lexical Resource For et.al NET 3.0 relative improvement of 21.96% Sentiment Analyses Of for the ranking by negativity Option Mining Twitter As A Corpus Opinion Mining  Large data set 6 Alexander Pak et.al For Sentiment Analyses And Sentiment  Improved needed increasing the And Potation Mining Analysis amount of the training data. Functional Analysis Of RojalinaPriyadershini Artificial Neural Back Propagation  Highly effective tool for classification. 7 et.al Network For Dataset Analysis  Combination of trail, learned Classification  while using the web mining for social Analyses Of Social Medias for analysis, data sampling Networks Using The Web 8 E.Raju, Sravanthi is a huge issue. TechniqueMove Web miningTechnique Mining  Filtering the data is difficult.  Communities Overlapping.  Sentiment detection used in the Sentiment Analyses different applications such as G.Vinodhini, Sentiment 9 And Opinion Mining A identify and classified reviews, Rm.Chandrasekaran Classification Survey summarize the review and other real time methods and applications . Churn Analyses Of Decision Tree  Analyze the potential mix of users. Xi Ling, Wenjing Online Social Network Classification 10  To process huge total of data in a Yin, et.al Uses Using Mining And K Means timely manner , decision tree-based Technique Algorithm and a k- means algorithms are used. A Survey On Social Social Network  Link analysis and deep and dark 11 Rupam Some Network And Future Models Statically networking Is recent trends in social Trends And For Analysis networks Paridhi Jain, Identifying Users  Identify the spammers and nemo while 12 PonnurangamKumara Across Multiple Social Finding Nemo using this algorithm guru, et.al Networks  More efficient in information finding process Comparison Of Web Neha Mehata, Manta Conventional And  Fuzzy clustering are suitable for 13 documentClusteri Kathuria et.al Fuzzy Clustering handling the issues related to ng Techniques A Survey understandability of patterns, incomplete/noisy data,  mixed media in rank and human

110

Advances in Engineering Research (AER), volume 142

communication, and can provide approximate data Neuro Fuzzy Approach Noro Fuzzy To Data Clustering A 14 FarhatAroochi Clustering  automatically generate the patterns Framework For Analysis Analyses Naïve Bayesian Atomic Classification Mita , K.Dalal, Model Artificial  it’s an domain dependent 15 Of Unstructured Blog Mukesh et.al Neutral Network Structure Text  Unstructured blog text classification Model A Survey On Using Datamining Techniques  with solid basis in graph theory , 16 G.Nandi,A.Das Graph Theory For Online Social analysis of data is done Network Analyses Semi supervised  To identify opinionated sentences Learning Based beginning unstructured consumer Mita K Dalal, OptimizedOpinionSum Semi Supervised 17 reviews with great success and Mukesh et.al marization And Approach categorize their orientation with Classification For acceptable accuracy Online Product Review  New robotic mechanisms are needed to help in drilling from downward to Critical Review On Sentiment top micro posts. 18 Asad Bukhari, et.al Sentiment Analyses Analysis Technique Technique  Improves the sentiment analysis classifiers that are able to enhance and analyze the user sentiments  By combining social network analysis Empirical Analysis with content mining using Technique For Data Amalgam approach would be more 19 S.G.S Fernando Mining Technique For Markov Model functional . Social Network  The statistical methods like Markova Websites models can be adopted to resolve the temporal behavior of web data.  Identify and Verdict communities in social networks structure, penetrating patterns in social Critical Analyses Of Sanjeev Dhawan, Web mining networks and investigative 20 Social Networks With Kulvinder Sing, et.al Technique overlapping communities. Web Data Mining  online social networking websites like online photo albums, comments and blogs.  Sentimental analysis includes Emotion Mining Sentiment managing customer dealings, Sanjeev Dhawan, 21 Technique In Social Analysis emotions of human ,information Kulvinder Singh, et.al Networking Sites Technique retrieval, natural text-to-speech system, social and literary analysis. A Study Of Data Knowledge Based  Focus to find global structural patterns. 22 M.Vedanayaki Mining And Social Network Analysis Network Analysis  tedious to collect data

111 Advances in Engineering Research (AER), volume 142

 Focus and analyze the extracted Sentiment MdAnsarulHaque, Sentiment Analyses By opinions from posted in social 23 Analysis TamjidRahaman Using Fuzzy Logic networks like facebook and twitter Technique etc.. A Survey Of Mariam Adedoyin- Datamining Technique  Conducted on social network analysis 24 Tram Olewe, et.al For Social Network using the tram technique. Analyses A Survey And Comparison Of Sentiment  using the sentimental analysis Anais Collomb, 25 Sentiment Analyses Analysis techniques , models target a simple CrinaCostea, et.al Methods For Technique global classification. Reputation Evaluation  Neuro and fuzzy logic approaches Clustering In Defaming Neuro And Fuzzy 26 Meenu Sharma helps to improve the clustering A Breve Review Logic Approaches review  Soft computing methodologies- solve Soft Computing Y.K. Soft Computing data mining problems. 27 Technique And Its MathurAbhayaNand Techniques Impact In Data Mining  Discover the no pattern in the large databases. Esmaeil Datamining Technique  Dynamic behavior of data. 28 FakhimiGheshlagh For Web Mining : A Web Mining et.al Review  Hybrid approach The Application Of Zahra 29 Link miningIn Social Link Mining  links among the data instances ZamaniAlavijeh, et.al Network Analysis  Used to extract relevant information Web Content Mining : from many websites likeonline Pooja Rohill, Ochin 30 A Implementation In Web Mining shopping sites. Sharma Social Websites  Data is extracted as per the user’s criteria using web mining  for text classification SVM algorithm A Survey In Text ShilpaRadhakrishana Machine Learning is mostly used. 31 Filtering In Online n Based Approach Social Networks  SVM algorithm has many features in social media analysis. Filter And Gauri Joshi, Classification Of User Naïve Bayes 32 Mr.SamadhanSonawa And Media Data Using MultilevelClassifi  To examining social medium statistics. ne Metric And er Algorithm NativeBased Method Exploring Different  An Exploratory research is carried by Aspects Of Social comparing adjacent matrices and Hilal Ahmed Datamining 33 Network Analyses lists with incidence matrices using Khanday, et.al Technique Using Web Mining graphs obtained from online social Techniques networks as data input Datamining Of Social K-Means  SVM is not suitable for mining using 34 Pooja Sikka Networks Using Clustering Based large data sets Clustering Based SVM SVM

112

Advances in Engineering Research (AER), volume 142

Classifying Shot Tecta Text-  Challenge in the classification of text FarisKateb, 35 In Social Media: Classification in social media is that the data is JugalKalita Twitter As Case Study Technique streamed Classification Of Adaptive Neural  learns to distinguish the required and G.Basiashavili, Application Of Multilayer Neutral 36 redundant and unnecessary e- mails T.Bliadze, et.al Adaptive Neural Technique from one another Networks For The Filtration Of Spam Opinion Mining RetuMevali,AjithSing 37 Techniques On Social OpinionMiming  take out precious thing of knowledge h, et.al Media Data  At the same time it permitted one Text Categorization comment to fall into multiple Naïve Bayes Using Data Mining categories. 38 `RemyaRs, SmithaEs Multilevel Technique On Social Classification  learning analytics, instructive data Media Data mining, and education related data mining are the benefits.  when the textual information is not Online Social Network Naïve Bayes Text 39 KanikaMathur structured as per the grammatical Mining Classification gathering, it becomes tough. Advanced Neuro- Fuzzy Approach For Hybrid Noro  Hybrid Noro Fuzzy helps to solve 40 Rajeev Mathur Social Media Mining Fuzzy different data mining problems Methods Using Cloud A Survey On Data R. Adikka,Dr .Shaik Mining Techniques For Datamining  Discuss the various data mining 41 et.al Analysis Of Social Technique techniques. Network Application Of Artificial Neural Network In Social Thai Le Phillip Pardo, 42 Media Data Analysis: A Anson Helps to the social analysis et.al Case Of Lodging Business In Philadelphia Sentiment Analysis Of K-Nearest  Give better results LopamudraDey, Review Datasets Using 43 Neighbor And Sanjay Chakroborty Naïve Bayes And K-  For hotel reviews these algorithms not ‘ Native Bayes NN Classifier suitable  Sentiment analysis of product reviews A New Way For Semi MuktaPatkar, Pooja gives optimistic and unenthusiastic Supervised Learning Semi Supervised 44 Pawar Money Sing, reviews Based On Data Mining Approach et.al For Product Reviews  Gives unbiased and constructive opinion Analytical  To identify the specified urn structure Dr.S.P.Victor, Web Structure 45 Implementation Of content analysis . Mr.M.Xavier Rex Web Structure Mining Mining

Using Data Analysis In

113 Advances in Engineering Research (AER), volume 142

Educational Domain Association Rule Hemant Kumar Soni, Mining: A Data Association Rule  Displays and describes the association 46 Snjiv Sharma, et.al Profiling And Mining rule mining . Prospective Approach Research On Pattern  Using the knowledge of machine Analysis And Data Pattern Analyses Heling Jiang, An learning methodology and graph Classification And Data 47 Yang,Fengyun Yan , theory to analyze the organizational Methodology For Data Classification et.al structure of network graphical Mining And Metrology pattern Knowledge Discovery An Efficient Data Mining Method To  For an item sets, direct frequency Saravanan Suba And 48 Find Frequent Item Sets TR-FCTM count and total frequency count are T.Chistophar In Large Database found Using TR- FCT V.A  Extraction of number of pages is Semantic Web Mining Semantic Web 49 ChakkarwarAmurta minimized by ranking technique Using RDF Data Mining A Joshi A Review On Web Kuldeep Sing Usage Mining For Web Clustering 50  Find hidden information. Rathore, et.al Personalization Using Technique Clustering Techniques Research Issues And Santhosh C.Pawar,  Semi and unstructured data in 51 Future Directions In Web Mining et.al challenging task Web Mining: A Survey

III. CONCLUSION [6] Marcelo Maia, et.al “Identifying User Behavior in Online Social Networks”, SocialNets’08, April 1, 2008 This research work provides a lot of present appraisal and [7] Ai Ho, Abdou Maiga, et.al “Privacy Protection Issues in Social updates of social media network analysis .In this Literature Networking Sites 2009 works are reviewed supported dissimilar aspects of social [8] L. Dey et.al “Opinion mining from noisy text data,”2009 networks. This work helps to studies the relevancy of the [9] Mohammad Al-Fayoumi, Soumya Banerjee, Jr.,et.al “Analysis of Social Network Using Clever Ant Colony Metaphor”, PROCEEDINGS OF techniques and idea of web mining for social networks WASET 2009 analysis, and reviews the connected literature concerning web [10] David Ediger Karl Jiang et.al“Massive Social Network Analysis:Mining mining and social networks. Data mining have several Twitter for Social Good”, ICPP 2010 challenges during this analysis field to be resolve with [11] Rojalina Priyadarshini; “Functional Analysis of Artificial Neural Network for Dataset Classification IJCCT August 2010. improvement. [12] Selman Bozkır1 , et.al Identification of User Patterns in Social Networks by Data Mining Techniques: IMCW 2010. REFERENCES [13] Alexander Pak et.al “Twitter as a Corpus for Sentiment Analysis and Opinion mining”, Proceedings of the LREC 2010. [1] Sankar K. Pal, et.al “Web Mining in Soft Computing Framework:, [14] S. Baccianella, et.al enhanced lexical resource for sentiment analysis and IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 13, NO. opinion mining,” in Proceedings of LREC ’10 2010 5,SEPTEMBER 2002. [15] Sitaram Asur et.al “ Predicting the Future With Social Media” Mar [2] Tomoyuki NANNO, et.al “Automatically Collecting,Monitoring, and 2010. Mining Japanese Weblogs”,. [16] Kuan-Yu Lin et.al, “Why people use social networking sites: An [3] Ralph Gross et.al “Information Revelation and Privacy in Online Social empirical study integrating network externalities and motivation Networks ACM Workshop on Privacy in the Electronic Society theory”,Computers in Human Behavior 27 (2011). (WPES), 2005. [17] M. Taboada et.al “Lexicon-based methods for sentiment analysis .IEEE [4] Huan Liu et.al “Toward Integrating Feature Selection Algorithms for 2011 Classification and Clustering”,IEEE Transactions 2005. [18] E.Raju K.Sravanthi, “Analysis of Social Networks Using the Techniques [5] Andrea Esuli_ et.al, “SENTIWORDNET: A Publicly Available Lexical of Web Mining” IJARCCSE Volume 2, Issue 10, October 2012. Resource for Opinion Mining”, Proceedings of the 5th Conference, [19] G. Vinodhini et.al, “Sentiment Analysis and Opinion Mining: A 2006. Survey”, IJARCSSE, Volume2, Issue 6, June 2012.

114

Advances in Engineering Research (AER), volume 142

[20] Fatemeh Khoshnood et.al “Designing A Recommender System Based [54] Ritu,Ajit Singh et.al, “Opinion Mining Techniques on Social Media On Social Networks And Location Based Services”, IJMIT 2012. Data”, IJCA 2 015. [21] Xi Long, Wenjing Yin, et.al “Churn Analysis of Online Social Network [55] Remya R S, Smitha E S, “Text Categorization using Data Mining Users Using Data Mining Techniques”, IJEMS 2012. Technique on Social Media Data”, [57] (IJARET Dec. 2015) [22] Simona Vinerean et.al“The Effects of Social Media Marketing on [56] saravanan et.al“An Efficient Data Mining Method To Find Frequent Online Consumer Behavior”,IJBM Vol. 8, No. 14; 2013 ISSN 1833- Item Sets In Large Database Using TR- FCTM ICTACT 2016. 3850. [57] V. A. Chakkarwar et.al “Semantic Web Mining using RDF Data”, IJCA [23] Rupam Some, “A Survey on Social Network Analysis and its Future Volume 133 – No.10, January 2016. Trends”, (IJARCCE) Vol. 2, Issue 6, June 2013. [58] Kuldeep Singh Rathore, et.al, “A Review on Web Usage Mining For [24] M. K. Dalalet.al, “Automatic classification of unstructured blog text,” Web Personalization Using Clustering Techniques”, IJARCCSER IJILSA, pp. 108–114, 2013. Volume 6, Issue 5, May 2016. [25] Paridhi Jainyet.al “Identifying Users across Multiple Online Social [59] Santosh C et.al “Research Issues and Future Directions in Web Mining: Networks”,IWWW 2013 A Survey”, IJCA 2016. [26] Farhat Roohi, “Neuro Fuzzy Approach To Data Clustering: A Framework For Analysis”, EJS ISSN: 1857 – 7881 . [27] G Nandi, A Das, “A Survey on Using Data Mining Techniques for Online Social Network Analysis”, 2013. [28] Mita K. Dalal et.al “Semisupervised Learning Based Opinion Summarization and Classification for Online Product Reviews”, Applied Computational Intelligence and Soft Computing Volume 2013. [29] S.G.S Fernando, “Empirical Analysis of Data Mining Techniques for Social Network Websites”, COMPUSOFT, IACT 2013 [30] Anaïs Collomb , et.al , “A Study and Comparison of Sentiment Analysis Methods for [31] Reputation Evaluation ,2014. [32] Sanjeev Dhawan et.al “Critical Analysis of with Web Data Mining”, IJITKM Special Issue (ICFTEM2014) May 2014 [33] M. Vedanayaki, “A Study of Data Mining and Social Network Analysis”,Indian Journal of Science and Technology, 2014. [34] Mariam Adedoyin-Olowe et.al “A Survey of Data Mining Techniques for Social Network Analysis”, JMDH , 2014. [35] Anu Sharma Dr. M.K Sharma et.al “CRITICAL REVIEW OF SENTIMENT ANALYSIS TECHNIQUES”, Proceeding of the ICAICS 2014 [36] Meenu Sharma, “Clustering In Data Mining : A Brief Review”, (IJCEM) Volume 1, Issue 5,August 2014. [37] Md. Ansarul Haque1, et.al “SENTIMENT ANALYSIS BY USING FUZZY LOGIC”, (IJCSEIT), 2014. [38] Y. K. Mathur, et.al “Soft Computing Techniques and its Impact in Data Mining – IJETAE August 2014). [39] Sanjeev Dhawan, et.al“Emotion Mining Techniques in Social Networking Sites”, IJICT (2014). [40] Haseena Rahmath P “Fuzzy based Sentiment Analysis of Online Product Reviews using Machine Learning Techniques IJCA 2015 [41] Esmaeil Fakhimi et.al“Data Mining Techniques for Web Mining: A Review 2015 [42] Zahra Zamani et.al “The Application of Link Mining in Social Network Analysis”, ACSIJ May 2015. [43] Thabit Zatari, “Data Mining in Social Media”,International Journal of Scientific & Engineering Research, Volume 6, Issue 7, July-2015. [44] Zohreh Madhoushi, et.al“Sentiment Analysis Techniques in Recent Works”, (SAI), 2015. IEEE, 2015. [45] G. Besiashvili, et.al“Application of Adaptive Neural Networks for the Filtration of Spam”, EPiC Series in Computer Science Volume 36, 2015, Pages 42–50 GCAI 2015. [46] Smita Bhanap, et.al“Data Mining for Business Intelligence in Social Network: A survey”, December 2015. [47] Literature Review and Challenges of Data Mining Techniques for Social 2015 [48] Pooja Rohilla, Ochin Sharma, “Web Content Mining: An Implementation on Social Websites”, IJARCCE, July 2015. [49] Shilpa Radhakrishnan, “A Survey on Text Filtering in Online Social Networks”, (IJCSIT) 2015 [50] Gauri Joshiet.al “Filtering and Classification Of User Based On Social Media Data Using Memetic and Naïve Bayes Methods”,(IJCSIT) 2015 [51] Hilal Ahmad Khanday, et.al“Exploring Different Aspects of Social Network Analysis Using Web Mining Techniques”, (IJETTCS) April 2015. [52] Pooja Sikka, “Data Mining Of Social Networks Using Clustering BasedSvm”, Ijrrest APRIL2015. [53] Faris Kateb,Jugal Kalita, “Classifying Short Text in Social Media: Twitter as Case Study”, IJCA February 2015.

115