Conceptual Framework of an Intelligent Decision Support System for Smart City Disaster Management

applied sciences

Article Conceptual Framework of an Intelligent Decision Support System for Smart City Disaster Management

Daekyo Jung 1, Vu Tran Tuan 1, Dai Quoc Tran 2 , Minsoo Park 2 and Seunghee Park 2,* 1 Department of Convergence Engineering for future City, Sungkyunkwan University, Suwon 16419, Korea; [email protected] (D.J.); [email protected] (V.T.T.) 2 School of Civil, Architectural Engineering & Landscape Architecture, Sungkyunkwan University, Suwon 16419, Korea; [email protected] (D.Q.T.); [email protected] (M.P.) * Correspondence: [email protected]

Received: 28 November 2019; Accepted: 13 January 2020; Published: 17 January 2020

Abstract: In order to protect human lives and infrastructure, as well as to minimize the risk of damage, it is important to predict and respond to natural disasters in advance. However, currently, the standardized disaster response system in South Korea still needs further advancement, and the response phase systems need to be improved to ensure that they are properly equipped to cope with natural disasters. Existing studies on intelligent disaster management systems (IDSSs) in South Korea have focused only on storms, floods, and earthquakes, and they have not used past data. This research proposes a new conceptual framework of an IDSS for disaster management, with particular attention paid to wildfires and cold/heat waves. The IDSS uses big data collected from open application programming interface (API) and artificial intelligence (AI) algorithms to help decision-makers make faster and more accurate decisions. In addition, a simple example of the use of a convolutional neural network (CNN) to detect fire in surveillance video has been developed, which can be used for automatic fire detection and provide an appropriate response. The system will also consider connecting to open source intelligence (OSINT) to identify vulnerabilities, mitigate risks, and develop more robust security policies than those currently in place to prevent cyber-attacks.

Keywords: decision support system; big data; artiﬁcial intelligence; Internet of Things; disaster management

1. Introduction Urban climate change is an important research issue that must be investigated using multidisciplinary approaches that include both engineering and socio-environmental sciences [1]. An eﬀective approach to strengthen disaster reduction strategies is to build a decision support system (DSS) for each region. A DSS is based on the correlation between the infrastructure, industries, and related communities in a region. Information on the technical, social, and economic properties of a region is collected to establish an eﬀective natural disaster prevention strategy. By analyzing the collected data, we can easily identify the vulnerabilities. Disaster-preparedness strategies for all citizens will make it easier to prepare and respond to real disasters.

Decision Support System A DSS is an information system that supports decision-making by selecting the best option by developing and comparing multiple alternatives to solve various issues. DSSs have a long history of development. From the late 1950s to the early 1960s at the Carnegie Institute of Technology, Nutt studied the theory of decision-making at an organization [2]. In the 1960s, MIT conducted technical research related to reciprocating computer systems. Subsequent studies by Pomerol and Adam

Appl. Sci. 2020, 10, 666; doi:10.3390/app10020666 www.mdpi.com/journal/applsci Appl. Sci. 2020, 10, 666 2 of 13

Appl.Sci, 2019, 11, x; doi: 2 of 13 contributed to the development of the DSS concept [3]. In 1981, Bonczek, Holsapple, and Whinston contributedestablished to a DSSthe frameworkdevelopment in of a book the DSS entitled concept “Foundations [3]. In 1981, of Decision Bonczek, Support Holsapple, Systems and”[4 ].Whinston Similarly, establishedin 1988, Turban a DSS et framework al. [5] proposed in a abook system entitled to solve “Foundations semistructured of Decision and unstructured Support Systems problems.” [4]. Similarly,This type in of 1988, effective Turban decision-making et al. [5] proposed by combining a system the to information solve semistructured gained from and data unstructured processing, problems.learning, andThis decision-makingtype of effective experiencedecision-making is known by combining as “intelligence the information DSS.” Today, gained DSSs arefrom proven data processing,to be an eff learning,ective support and decision- system inmaking businesses. experience In addition, is known DSSs as “intelligence have been the DSS foundation.” Today, forDSS thes aredevelopment proven to ofbe numerous an effective hypotheses support and system tools, suchin businesses. as artificial In intelligence addition, (AI),DSSs human–computer have been the foundationinteraction, for simulation the development methods, of software numerous engineering, hypotheses and and telecommunication tools, such as artificial for applications intelligence in (AI),DSSs. Thehuman use– ofcomputer DSSs has growninteraction, rapidly simulation since the 1950s, methods, particularly software with the engineering, aid of effective and data telecommunicationanalysis tools, enabling for applications improved decision-making in DSSs. The use using of DSSs data andhas information.grown rapidly Several since attempts the 1950s, are particularlybeing made with by scientists the aid of to effective improve data the e ffanalysisectiveness tools, of decision-makingenabling improved by decision-making combining technologies using datafrom and related information. fields. Several attempts are being made by scientists to improve the effectiveness of decision-makingThe components by combining of a DSS technologies include data from management, related fields. model management, user interface, knowledgeThe components management, of a andDSS users,include as data presented management, in Figure model1. A DSSmanagement, is an interactive user interface, system knowledgethat allows management, decision-makers and to users, easily as analyze presented and evaluate in Figure decision 1. A DSS models is an and interactive to process system the data that to allowssolve complex,decision-makers unstructured, to easily and analyze nonrepetitive and evaluate decision-making decision models tasks. Theand decision-makersto process the data access to solvethe system complex, via unstructured, the interface and nonrepetitive management decision-making components and tasks. extract The various decision-makers types of data access and theinformation system via from the the interface database and model management according components to their requirements. and extract various types of data and

Other Computer- Data : External Based Systems and Internal

Data Model Management Management

Knowledge Management

User Interface

Manager (User)

Figure 1. Decision support system components. Figure 1. Decision support system components. (a) Data Management: The data management component comprises a database that stores various data for decision-making as well as a database management system (DBMS). The function of the a) Data Management: The data management component comprises a database that stores DBMS in a DSS is to store and provide the data for the decision-making. various data for decision-making as well as a database management system (DBMS). The (b) Modelfunction Management: of the DBMSThe in model a DSS managementis to store and component provide the comprises data for the a modeldecision base-making. that stores b)variousModel modelsManagement: necessary The formodel the m decision-makinganagement component as well ascomprises a model basea model management base that systemstores (MBMS).various Particularly, models necessary the MBMS for playsthe decision a key role-making in decision as well support as a model by providing base management functions to develop,system modify,(MBMS). and Particularly control the, the models MBMS required plays a for key the role decision-making. in decision support by providing (c) Userfunctions Interface: to Thedevelop, user interfacemodify, and is the control module the systemmodels providing required for an the interface decision between-making. the user c)andUser the Interface: system forThe importing user interface and exporting is the module data and system performing providing various an interface analytical between procedures. the It isuser also and known the as system the dialogue for importing generation and and exporting management data softwareand performing (DGMS) various because analytical it provides procedures. It is also known as the dialogue generation and management software (DGMS) because it provides user-friendly and dialogue functions that are easy to understand and use, via menus or graphics processing formats. Appl. Sci. 2020, 10, 666 3 of 13

user-friendly and dialogue functions that are easy to understand and use, via menus or graphics processing formats. (d) Knowledge Management: This module provides quantitative information about the relationships between complex data. Knowledge management provides decision-makers with knowledge and alternative solutions for problem solving. It also signals to the decision-makers if there is a diﬀerence between the predicted result and the actual result. (e) Users: The users who use a DSS are primarily the managers who are responsible for important business decisions. They choose the most appropriate model from the model base, enter the necessary data from the database or import them directly into the model, and then evaluate and analyze the options to determine the best alternative.

This paper is organized as follows. Section2 presents a literature review about the applications of DSSs, AI, and big data for disaster management. Section3 discusses the methodology regarding the conceptual framework of an intelligent DSS (IDSS). It also introduces the recent problems of heatwaves, cold waves, and wildﬁres, and then proposes an AI IDSS system. Finally, Section4 presents the results and future research.

2. Literature Review

2.1. DSS for Disaster Management Although the use of DSSs to support disaster recovery efforts has provided tremendous benefits, these systems have certain disadvantages. One of the disadvantages is the inability to freely and rapidly transfer data between individuals and organizations. This is because the systems and technologies of different organizations are different or incompatible. Nonetheless, DSSs have overcome these disadvantages and are widely used by managers. Rajabifard et al. [6] used an intelligent disaster DSS (IDDSS) as a platform to integrate road, traffic, geographic, economic, and meteorological data. IDDSSs are used in the management of road networks during floods. To prevent hazardous traffic scenarios, they provide the law enforcement with the exact locations to establish traffic management points (TMPs) during an emergency. In 2011, Ishak et al. [7] created a conceptual model of a smart DSS for reservoir operations in case of emergencies such as heavy rainfall. This model can help reservoir operators make accurate decisions for releasing reservoir water so that there is sufficient space for the released water, to avoid local flooding. Moreover, AI has also been integrated into DSSs to increase the decision-making efficiency. Dijkstra’s algorithm can find the shortest path between two points and has several applications in various areas. This algorithm has been widely employed in forest fire simulations [8] and in improving the efficiency of route planning [9]. In 2011, Akay et al. [10] improved this algorithm using the Geographic Information System (GIS) to assist firefighters in determining the fastest and safest access routes. This system requires numerous spatial databases, including those of road systems and land. Barrier systems have also been established to simulate the scenario of banned roads; hence, these systems are not only used to determine the fastest route but also to assist firefighters in identifying unforeseen scenarios and determining safe and reliable routes. To deal with disasters, we require an effective disaster management system. However, disaster management systems depend on various data and information and knowledge of previous disasters. Therefore, coordination between the relevant agencies is necessary to integrate the data effectively. A DSS is unable to prevent all catastrophic damage; however, it can mitigate the potential risks by developing early warning strategies and preparing appropriate options. Particularly in the context of disaster management [11], the process of developing intelligent systems focuses on the methods of transferring expertise from human experts to computers and implementing theoretical models based on the transferred knowledge. The model proposed by Turban (Figure2) included the five main activities of DSS components [12]. Within the scope of AI, there are several methods for solving interdependent problems. Examples of alternative methods include Bayesian networks, which approximate reasoning (fuzzy logic), and metadata (artificial neural networks, machine learning, genetic algorithms, and Appl. Sci. 2020, 10, 666 4 of 13 swarm intelligence), which involves the use of a hybrid approach combining the fuzzy IF–THEN rules andAppl.Sci, optimization 2019, 11, x; doi: techniques to collect and present inaccurate information [13]. 4 of 13

Problem

Knowledge Acquisition Raw Knowledge

Knowledge Codified Representation Knowledge

Validated Knowledge Validation Knowledge

Metaknowledge Inferencing

Advice + Explanation Revision Feedback Loop

Solution

Figure 2. The process of knowledge engineering (adapted). Figure 2. The process of knowledge engineering (adapted). 2.2. AI for Disaster Management 2.2. AIAI for is usedDisaster in various Management applications, such as customer service, transactions, and healthcare. Recently, researchersAI is used have in found various that applicationsAI can be employed, such as to customer predict disasters. service, transactions, With numerous and high-quality healthcare. datasets,Recently, AI researchers can predict have the found occurrence that AIof natural can be disasters,employed which to predict canassist disasters. in saving Withthe numerous lives of thousandshigh-quality of datasets, people and AI incan preventing predict the financial occurrence losses. of Naturalnatural disasters, disasters thatwhich can can be predictedassist in saving by AI includethe lives earthquakes, of thousands floods, of people storms, and forest in preventing fires, and financ volcanicial eruptions. losses. Natural In the disasters field of earthquakethat can be prediction,predicted by researchers AI include are earthquakes, collecting vastfloods, amounts storms, of forest seismic fires, data and for volcanic analysis eruptions. using deep-learning In the field systems.of earthquake Researchers prediction, use researchers the data collected are collecting from past vast earthquakes amounts of toseismic make data improved for analysis predictions using fordeep future-learning earthquakes. systems. AlarifiResearchers et al. [use14] the proposed data collected an AI prediction from past system earthquakes based to on make artificial improved neural networks,predictions which for future could beearthquakes. used to predict Alarifi the et magnitude al. [14] proposed of an earthquake. an AI prediction AI can also system employ based seismic on dataartificial to analyze neural the networks magnitude, which and patterncould be of anused earthquake, to predict as the well magnitu as to aidde in of predicting an earthquake. the occurrence AI can ofalso an employ earthquake. seismic Goymann data to etanalyze al. [15 ]the presented magnitude data and for thepattern establishment, of an earthquake training,, as and well evaluation as to aid ofin apredicting neural network the occurrence to detect of flood an earthquake. and water levels. Goymann AI-based et al. systems[15] presented can learn data rainfall for the records, establishment, climate records,training, and and flood evaluation simulation of a neural tests and network can predict to detect floods flood better and than water traditional levels. AI systems.-based systems can learn rainfall records, climate records, and flood simulation tests and can predict floods better than 2.3. Big Data for Disaster Management traditional systems. Big data result from the large amount of data generated by the explosive growth of information technology2.3. Big Data and for D communications,isaster Management as well as by the rapid spread of smart devices [16,17]. They have beenBig applied data inresult several from fields the large [18,19 amount]. Using of large data datasets, generated it isby possible the explosive to analyze growth past of disaster information data andtechnology develop and future communications scenarios for, theas well possibilities as by the of rapid disasters spread and of thesmart damage devices caused [16,17] by. They them have [20]. Bigbeen data applied contribute in several to the fields science [18,19] of decision-making. Using large datasets by developing, it is poss a technologyible to analyze that past canread disaster common data patternsand develop of user future behaviors scenarios and for predict the possibilities future steps of [21 disasters]. AI is defined and the as damage a computer caused program by them that [20] can. Big data contribute to the science of decision-making by developing a technology that can read common patterns of user behaviors and predict future steps [21]. AI is defined as a computer program that can think and process information like a human being. It is the science that enables computers to perform intellectual activities based on experience and knowledge. Existing data serve

Appl.Sci,Appl. 2019 Sci. 2020, 11,, x;10 ,doi: 666 5 of 135 of 13 as the criteria for core judgments, and an AI learning algorithm narrows the choices for think and process information like a human being. It is the science that enables computers to perform decision-making and provides feedback that is very similar to human judgment. For an optimal intellectual activities based on experience and knowledge. Existing data serve as the criteria for response to complex disasters, a large amount of data should be analyzed and used; hence, AI can be core judgments, and an AI learning algorithm narrows the choices for decision-making and provides usedfeedback to understand that is very large similar-scale to humansocial phenomena judgment. For [22] an. optimalBy big responsedata analysis to complex and disasters, simulation, a the selectionlarge amountof the required of data should actual be data analyzed can be and achieved used; hence, using AI the can judgment be used to of understand AI based large-scaleon the existing empiricalsocial phenomena data. [22]. By big data analysis and simulation, the selection of the required actual data can be achieved using the judgment of AI based on the existing empirical data. 2.4. Intelligent Decision Support Systems 2.4. Intelligent Decision Support Systems Using AI, researchers have attempted to improve the effectiveness of decision-making. Using AI, researchers have attempted to improve the effectiveness of decision-making. Integrated Integrated AI support systems or IDSSs have been created and used in various fields, such as AI support systems or IDSSs have been created and used in various fields, such as healthcare healthcare and commerce. and commerce. TheseThese systems systems use use AI AI tools tools to infer, to infer, learn, learn, memorize, memorize, plan, and plan analyze, and the data.analyze Simon the presented data. Simon presentfoured stages four in stages a decision-making in a decision process,-making which process, are displayed which are in displayed Figure3[ 23 in]. IntelligenceFigure 3 [23] is. the Intelligence first is thestage first of stage the study of the and study requires and the requires decision-makers the decision to develop-makers an understandingto develop an of understanding the problem and of the problemcollect and information collect information related to the related decision-making. to the decision Similarly,-making. the design Similarly, stage involves the design characterizing stage involves characterizingthe data with the important data with variables important based variables on decision based issues, on determining decision issues, the criteria determining for the decisions, the criteria for theand decisions, developing and decision developing models thatdecision can be models used to that evaluate can alternativebe used to decisions.evaluate Inalternative the third stage,decisions. In thethe third decision-makers stage, the decision review the-makers alternatives review and the choose alternatives the consolidated and choose decisions the consolidated that satisfy decisions the thatalternatives. satisfy the The alternatives. final stage, sometimes The final referred stage, to sometimes as evaluation, referred is when to the as decision-makers evaluation, is assesswhen the the consequences of a decision. decision-makers assess the consequences of a decision.

INTELLIGENCE • Observe and understand the problem • Develop a statement of the problem and decision • Acquire information relevant to the decision

DESIGN • Identify variables and criteria for the decision • Specify relationship between variables • Identify controllable and uncontrollable events • Develop decision models that can be used to evaluate alternatives • Develop alternatives

CHOICE • Evaluate decision alternatives • Make or recommend a decision that best meets the decision criteria

IMPLEMENTATION • Consider consequences of the decision • Gain confidence in the decision • Develop an implementation plan • Secure needed resources • Put implementation plan into action

Figure 3. Four decision-making stages Figure 3. Four decision-making stages

IDSSs utilize AI techniques to provide better support to decision-makers. AI tools, such as fuzzy logic, case-based reasoning, evolutionary computing, artificial neural networks, and intelligent agents, when combined with a DSS provide powerful methods for solving difficult problems that are frequently real-time and involve large amounts of distributed data [24]. IDSSs utilize some of the below-mentioned tools: o Fuzzy logic for intelligent decision support;

Appl. Sci. 2020, 10, 666 6 of 13

IDSSs utilize AI techniques to provide better support to decision-makers. AI tools, such as fuzzy logic, case-based reasoning, evolutionary computing, artiﬁcial neural networks, and intelligent agents, when combined with a DSS provide powerful methods for solving diﬃcult problems that are frequently real-time and involve large amounts of distributed data [24]. IDSSs utilize some of the below-mentioned tools:

Fuzzy logic for intelligent decision support; # Expert systems for intelligent decision support; # Evolutionary computing for intelligent decision support; # Intelligent agents for intelligent decision support. # Disaster response is a continuous decision-making process based on past experiences, because a disaster scenario is recognized as a disaster until it ends. In the past, because of the limitations regarding information and data-collection technology, the decision-making process had various constraints and relied on the subjective experiences of the decision-makers. Recently, however, information technology has developed considerably and the amount of information that can be used to make decisions has increased sharply. The information processing ability of humans is limited, and responding to disasters requires timely decisions. The benefits of combining a DSS and AI to create an IDSS in disaster management are clear. However, the use of information of past disasters to analyze and make predictions has not sufficiently been investigated. Therefore, this study aims to contribute to this growing area of research by exploring a conceptual framework that supports intelligent decision-making using big data to help decision-makers make more rapid and more accurate decisions. Recently, a series of innovative solutions based on free and open-source software has been developed [25–27]. Open source information is data collected from publicly available sources and the system data. In this study, the proposed system also collected data from various open sources. With the growth of big data and online social networks, increasing amounts of data are available online, which creates opportunities for criminals—particularly adversaries preparing for a cyberattack [28]. Thus, an IDSS can become vulnerable to cyberattacks. Consequently, the scientific community, companies, practitioners, and governments worldwide are searching for solutions to mitigate risks [29]. Hayes and Cappa [30] introduced open source intelligence (OSINT) to provide tactical and strategic recommendations for organizations to prevent cyberattacks by identifying vulnerabilities, mitigating risks, and formulating robust security policies.

3. Methodology Heat waves, cold waves, and wildfires are the main concerns for the South Korean government, because of the difficulty in predicting and responding to these events. Therefore, the proposed AI IDSS focuses on the aforementioned phenomena. In the next two subsections, the issues and scenarios regarding heat waves, cold waves, and wildfire phenomena in South Korea are briefly introduced. Based on these issues, an AI IDSS with a multifunction AI algorithm is presented.

3.1. Issues and Scenarios of Heatwave/Cold Wave Disasters The fund for disaster support money paid to the families of those who are killed by heat is limited between 5000 and 10,000 dollars (in the case of household heads) based on the South Korea disaster relief regulations and costs (DRRCs) for disaster recovery. However, no system or technology has been established to objectively determine the causes of deaths during the summer in South Korea. In addition, there have been numerous diﬃculties in diagnosing the causes of deaths owing to victim interventions and ethical issues that may have arisen. Furthermore, while heatwave damage occurs in almost all areas during the summer, the extent of the damage varies depending on the local scenarios and personal health conditions, which can lead to legal repercussions. Appl. Sci. 2020, 10, 666 7 of 13

A domestic cold wave alarm is issued based on three criteria. It is issued during the winter (October to April), when the minimum temperature the following morning is expected to drop by more than 10 ◦C compared to that of the previous day and is below 3 ◦C, or when the temperature is lower than the average winter temperature values of previous years by 3 ◦C. The warning is also issued when the minimum morning temperature is expected to be below 12 ◦C for two days or more, or when there is a significant loss of life and property due to a sudden fall in the temperature. If there is a specific list of victims or temperature damages, it is possible to set an absolute standard for facility damages or loss of life. However, as such a database has not yet been established, cold weather management will need to be implemented and systematically managed by generating the associated database. There is no big data AI study case for heat and cold waves. In 2018, the Ministry of Public Administration and Security identified heat and cold waves as new types of disasters. Currently, there is no specific management standard to deal with crises caused by heat and cold waves in Korea. It is also necessary to improve the speed and accuracy of the system that predicts heat and cold waves. For example, it is possible to predict pre-heat phenomena two days in advance; however, in developed regions such as Europe or the U.S., this can be done three to five days in advance. Through this study, by collecting data to prevent and be prepared for disasters, an AI-based support system was developed to enable more rapid and more accurate decision-making than is currently achievable.

3.2. Issues and Situations Of Forest Fire Disaster When considering information on the direction and speed of a spreading fire, a possibility of widespread fire is found to exist. Accordingly, a rapid plan to protect nearby houses and evacuate people is required. However, a gap remains in relaying the wildfire information from the fire helicopter to the ground firefighter commander. Even if a firefighter commander decides that an emergency defense is needed, it is still necessary to go through the integrated fire protection procedure to adopt the necessary measures. In the process, damages are frequently increased because of the inappropriate propagation of information or lack of content transformation. Therefore, it is necessary to improve the information sharing system for handling forest fires.

3.3. A System Using Big Data and AI for Intelligent Decision Support The level of success of a DSS is assessed using two parameters. The first parameter is the data input rate and output rate, which represent the information that is useful for the decision-making process. The second parameter is time. To improve the efficiency of the DSS in this study, a conceptual framework using big data and AI to enhance the above-mentioned parameters is proposed. The decision points of the standard manual for heat/cold wave and forest fire crisis management and the services provided by the decision-maker in the event of a disaster are summarized in Figure4, along with the big data system processes and big data analysis. The proposed big data process is composed of data sources, collection, storage, analysis, and presentation order, and is built on the basis of the data lake (DL) form. The data collection is performed based on the type of natural disaster that a given DSS is built to deal with. The choice of data source for the analysis is extremely important. If the collected data are not reliable, they will be difficult to process and discard. Extreme disasters, such as heat and cold waves, need to be analyzed, forecasted, and planned for in advance. By contrast, with forest fires, the post-disaster response includes population evacuation, which is more important than forecasting. Therefore, camera data should be analyzed to predict the speed and direction of the flames and to send information to the people in the affected areas. Appl. Sci. 2020, 10, 666 8 of 13 Appl.Sci, 2019, 11, x; doi: 8 of 13

Figure 4. System using big data and AI for intelligent decision support. Figure 4. System using big data and AI for intelligent decision support. To collect and store data, it is important to build a DL. A DL differs from a data warehouse (DW) in that theTo former collect contains and store both data, structured it is important and unstructured to build a DL. data. A DL Post differs data from collection, a data warehouse it undergoes (DW) the processin that of the extraction, former contains transformation, both structured and loading and unstructured (ETL); however, data. this Post process data collection, is time-consuming it undergoes and makesthe it process difficult of to arrive extraction, at a timely transformation, decision. Hence, and this loading system (ET splitsL); the however DL archive, this into process two parts. is Thetime first-consuming is a platform and that makes can storeit difficult the data to arrive in the at original a timely form decision. without Hence, going this through system the splits ETL process,the DL andarchive the second into two is theparts data. The analysis first is platforma platform that that undertakes can store the the data ETL in process. the original The collectedform without data includegoing real-timethrough the data ETL and process, batch data.and the Real-time second is data the data are structured analysis platform data, which that undertakes can be periodically the ETL collectedprocess and. The processed collected data by the include existing real DWs.-time Batchdata and data batch are thedata. raw Real and-time unstructured data are structured data that data need, towhich be used can for be analysis. periodically collected and processed by the existing DWs. Batch data are the raw and unstructured data that need to be used for analysis. 3.4. Data Collection and Storage 3.4. Data Collection and Storage Essentially, most structured data are collected by an open API. The data are provided in Comma-SeparatedEssentially, most Values structured (CSV) format data are on collected a public by data an open portal API. operated The data by are the provided Ministry in of AdministrationComma-Separated and Security. Values Because(CSV) format big data on systems a public use data data portalfrom various operated organizations, by the Ministry the range of ofAdministration the collected data and is Security. wide. However, Because only big data the data systems considered use data necessary from various for the organizations, analysis must the be used.range In of the the case collected of image data data is thatwide require. However real-time, only the processing, data considered satellite necessary image information for the analysis can be usedmust to be cover used a. wideIn the area.case of Along image with data satellitethat require images, real- Closed-Circuittime processing, Television satellite image (CCTV) information cameras cancan collect be used image to data cover anytime a wide and area. anywhere, Along with and satellite they can images, be used Closed to detect-Circuit a fire Television early. Most (CCTV CCTV) camerascameras in can South collect Korea image are useddata anytime for traffi andc or anywhere, crime prevention, and they and can their be us dataed to storage detect cyclea fire isearly. only 30Most days. CCTV Thus, cameras their application in South toKorea machine are used learning for traffic in the or disaster crime prevention, field is limited. and Thesetheir data elements storage are listedcycle in is Table only1 , 30 which days. describes Thus, their the types application of disaster, to machine data sources, learning and in organizations the disaster field responsible is limited. for These elements are listed in Table 1, which describes the types of disaster, data sources, and answering the questions of decision-makers and enabling them to make rapid and precise decisions. organizations responsible for answering the questions of decision-makers and enabling them to make rapid and precise decisions.

Table 1. Questions and data sources required for disaster response (Application Programming Interface (API), Closed-Circuit Television (CCTV), Unmanned Aerial Vehicle (UAV))

Disaster Data Source Question Type Extreme Statistical Geographic Information Current status of population and households Weather Service (Open API) in heat/cold wave expected areas

Appl. Sci. 2020, 10, 666 9 of 13

Table 1. Questions and data sources required for disaster response (Application Programming Interface (API), Closed-Circuit Television (CCTV), Unmanned Aerial Vehicle (UAV)).

Disaster Type Data Source Question Statistical Geographic Information Current status of population and households Extreme Weather Service (Open API) in heat/cold wave expected areas Statistical Geographic Information Status of land use and age of buildings in Extreme Weather Appl.Sci, 2019, 11, x; doi: Service (Open API) heat/cold wave expected areas9 of 13 Status of operation of hot/cold shelter in Extreme Weather Data Portal (Open API) Extreme Statistical Geographic Information Status of landheat use/cold and wave age of expected buildings areas in Weather Service (Open API) heat/cold wave expected areas Korea Centers for Disaster Control Cases of human damage from past ExtremeExtreme Weather Status of operation of hot/cold shelter in Dataand PreventionPortal (Open (Document) API) heat/cold waves Weather heat/cold wave expected areas ExtremeExtreme Weather KoreaKorea Centers Meteorological for Disaster Administration Control Cases Weatherof human status damage (temperature, from past heat/ precipitation,cold WildfireWeather and Prevention(Open (Document) API) wind, humidity,waves atmospheric pressure) Extreme Extreme Weather Korea Meteorological Weather status (temperature, precipitation, Weather CCTV, UAV, Satellite (Image, Video) Disaster image information Wildfire Administration (Open API) wind, humidity, atmospheric pressure) Wildfire Status of water supply in heat/cold wave Heat Wave K-Water (Open API) Extreme expected areas Weather CCTV, UAV, Satellite (Image, Video) Disaster image information WildfireWildfire Data Portal (Open API) Forest fire danger rating index Wildfire Data Portal (Open API)Status of water Past supply wildfire in h disastereat/cold caseswave Heat Wave K-Water (Open API) expected areas Korea Database on Protected Areas Status of cultural assets and hazardous Wildfire Wildfire Data Portal (CSV)(Open API) Forest facilitiesfire danger near rating the forestindex area Wildfire Data Portal (Open API) Past wildfire disaster cases Korea Database on Protected Areas Status of cultural assets and hazardous Wildfire 3.5. Machine Learning Based on a Data(CSV) Analysis Algorithm for AI Managementfacilities near the forest area The algorithm modeled for the machine learning data analysis is displayed in Figure5. It uses 3.5. Machine Learning Based on a Data Analysis Algorithm for AI Management convolutional neural network (CNN) and generative adversarial network (GAN) algorithms for natural disaster detectionThe algorithm and employs modeled reinforcementfor the machine learninglearning data (RL) analysis to predict is displayed wildfire in direction. Figure 5. It Auses CNN is a specificconvolutional type of artificial neural neural network network (CNN) thatand generative uses perceptron, adversarial a machine network learning(GAN) algorithms unit algorithm, for to natural disaster detection and employs reinforcement learning (RL) to predict wildfire direction. A analyze the data. A CNN can be applied to image processing, natural language processing, and other types of cognitive tasks.

Wildfire response: a proposed system (Closed-Circuit Television :CCTV)

Past Weather Sensor Satellite events data data data data Prepare Prepare Convolutional Generative Reinforcement Neural Adversarial learning Network Network

CCTV Trained real-time model Fire? No Store data (Detection) Detect Drone real-time Yes data

Trained Response model suggestion (Prediction) Response

Figure 5. Wildﬁre response: the proposed system diagram. Figure 5. Wildfire response: the proposed system diagram. Appl.Appl.Sci, Sci. 20192020,, 1110,, x; 666 doi: 1010 of of 13

The GAN algorithm, which is an unsupervised learning algorithm used in the model, utilizes imageThe data GAN both algorithm, from fires which and is anfrom unsupervised normal conditions learning algorithm for learning. used Real in the-time model, CCTV utilizes and imageUnmanned data bothAerial from Vehicle fires and(UAV from) data normal are then conditions fed into for and learning. trained Real-time in the models CCTV to and determine Unmanned the Aerialoccurrence Vehicle of (UAV) a fire. data If it are is thennot a fed forest into andfire, trained then the in theused models data willto determine be stored the and occurrence utilized of for a fire.subsequent If it is not training. a forest fire, In the then case the used of forest data will fires, be significant stored and utilizedinformation for subsequent about the training. weather In and the casereal- oftime forest satell fires,ite images significant are informationrequired. The about weather the weather data include and real-time the temperature, satellite images wind speed, are required. wind Thedirection, weather rainfall, data include and humidity. the temperature, At this point, wind speed,the model wind is direction, transformed rainfall, into and an humidity.enhanced Atmodel this point,that predicts the model the propagation is transformed speed into and an enhanced the direction model of the that fire predicts per unit the of propagation space. speed and the directionThe following of the fire perdiscussion unit of space.presents how the proposed algorithm can be used for fire detection and response.The following CNN discussions were first presents introduced how theby proposedLeCun et algorithmal. [31]. By can collective be used knowledge for fire detection from andthe response.training data, CNNs computers were first can introduced learn and by understand LeCun et al. a [struc31]. Byture collective based on knowledge a hierarchy from of concepts. the training A data,common computers CNN architecture can learn and contains understand various a structure types ofbased processing on a hierarchy layers of including concepts. convolution, A common CNNpooling, architecture and fully containsconnected various layers types[32]. The of processing optimization layers process including of a CNN convolution, is conducted pooling, by trial and fullyand error connected and guided layers by [32 ch].ecking The optimization the validation process set error of a [33] CNN. In is this conducted study, the by overall trial and architecture error and guidedand layers by checkingof FireNet the [33] validation were used. set errorThe FireNet [33]. In thisCNN study, architecture the overall contains architecture three convolutional and layers of FireNetlayers of [ 33sizes] were 64, used.128, and The 256, FireNet with CNNkernel architecture filters of sizes contains 5 × 5, 4 three × 4, convolutionaland 1 × 1, respectively. layers of sizesEach 64, 128, and 256, with kernel filters of sizes 5 5, 4 4, and 1 1, respectively. Each convolutional convolutional layer is followed by a max-pooling× layer× with a kernel× size of 3 × 3 and local response layer is followed by a max-pooling layer with a kernel size of 3 3 and local response normalization. normalization. This set of convolutional layers is followed by two× fully connected layers, each with This4096 setincoming of convolutional connections layers and a is hyperbolic followed by tangent two fully activation connected function. layers, A eachdropout with of 4096 0.5 is incoming applied connectionsacross these andtwo afully hyperbolic connected tangent layers activation during the function. training A to dropout offset the of 0.5 residual is applied over across-fitting. these Finally, two fullywe obtain connected a fully layers connected during layer the trainingwith two to incoming offset the residualconnections over-fitting. and a soft Finally,-max activation we obtain output a fully connected[33]. layer with two incoming connections and a soft-max activation output [33]. The experimentexperiment waswas performedperformed using using a a dataset dataset of of 9852 9852 images images from from Chenebert Chenebert et al.et [al.33 ],[33] and, and the testingthe test wasing was conducted conducted on 2931on 2931 images. images. The experiment was performed using an NVIDIA GeForce GTX 2080 Ti with 12 GB onboard memory andand aa deep deep learning learning framework framework [34 [34]] on on an an Intel Intel Xeon Xeon CPU CPU E3-1230 E3-1230 v6 with v6 with 32 GB 32 RAM. GB RAM. From theFrom results, the results it was, notedit was thatnoted FireNet that FireNet could predict could thepredict test datathe test with data an accuracywith an accuracy of 91.2%. of Figure 91.2%.6 displaysFigure 6 thedisplays grad classthe activationgrad class map activation (CAM) map results, (CAM elucidating) results,how elucidating the computer how thediff erentiated computer betweendifferentiated a fire between and a non-fire. a fire and This a demonstratesnon-fire. This thatdemonstrates this algorithm that this can algorithm be used for can automatic be used firefor detectionautomatic and fire candetection suggest and the can appropriate suggest the response. appropriate response.

Figure 6. FireNet model localizes the ﬁre area using class activation maps (CAMs), which highlight the Figure 6. FireNet model localizes the fire area using class activation maps (CAMs), which highlight areas of the image that are most important for making predictions. the areas of the image that are most important for making predictions. 4. Conclusions 4. Conclusion With climate change, natural disasters have increased in frequency and intensity compared to thoseWith in the climate past. Thus,change, in thisnatural study, disasters we proposed have increased a support in system frequency for disasterand intensity response compare decisionsd to usingthose bigin the data past. storage Thus, and in machinethis study, learning we proposed analysis. a Wesupport focused system on forest for disaster ﬁres and response hot/cold decisions disasters, using big data storage and machine learning analysis. We focused on forest fires and hot/cold

Appl. Sci. 2020, 10, 666 11 of 13 which cause significant loss of life and property annually. Current disaster area systems focusing on simple methods of information collection, efficient visualization, storage systems, and big data analysis are inadequate. In other countries, big data analysis conducted during the entire process of disaster management, including prevention, preparation, response, and recovery, mainly uses video data and social media. Based on previous studies, we developed a big data analysis algorithm to support decisions for heat wave disasters and proposed its application. In the field of disaster management, which requires prompt decisions, if the current forest fire detection system employs image recognition technology intelligently, preventive responses can reduce the damage. In addition, it can provide information that may be useful for decision-making, instead of yielding simple statistical information. Moreover, a simple example of using a CNN to detect fire in a surveillance video was introduced. It was evident that this algorithm could be used for automatic fire detection and provide an appropriate response. This conceptual framework can support decision-makers and enable them to make correct decisions that can save lives, reduce property damage, and assist in predicting disasters. In the future, the system and the algorithm will be tested with real events to gradually increase its accuracy by diversifying the input data and correcting the algorithms. We will also consider connecting the system with OSINT to identify vulnerabilities, mitigate risks, and formulate more robust security policies than the current ones to prevent cyberattacks.

Author Contributions: D.J. Developed Ideas and System Modeling; V.T.T. Methodology and Manuscript Writing; D.Q.T. Methodology and Edited the Paper; M.P. Project administration; S.P. Supervision and Funding Acquisition. All authors have read and agreed to the published version of the manuscript. Funding: This research was supported by a grant [2019-MOIS31-011] from the Fundamental Technology Development Program for Extreme Disaster Response funded by the Ministry of Interior and Safety [MOIS, Korea] and supported by the Korea Ministry of Land, Infrastructure and Transport [MOLIT] as an [Innovative Talent Education Program for Smart City]. Conﬂicts of Interest: The authors declared no conﬂict of interest.

Abbreviations The following abbreviations are used in this manuscript:

Decision Support System DSS Big Data BD Internet of Things IOT Database Management System DBMS Model Base Management System MBMS Dialogue Management and Management Software DGMS Intelligent Disaster Decision Support System IDDSS Traﬃc Management Points TMP Artiﬁcial Intelligence AI Intelligent Decision Support System IDSS Open Source Intelligence OSINT Disaster Relief Regulations and Costs DRRC Data Base DB Data Lakes DL Extraction, Transformation, and Loading ETL Social Networking Service SNS User Interface UI k-Nearest Neighbor KNN Convolutional Neural Network CNN Recurrent Neural Network RNN Multilayer Perceptron MLP Support Vector Machine SVM Generative Adversarial Network GAN Bidirectional Encoder Representations from Transformers BERT Appl. Sci. 2020, 10, 666 12 of 13

Data Warehouse DW Comma-Separated Values CSV Closed-Circuit Television CCTV Class Activation Maps CAMs Unmanned Aerial Vehicle UAV Application Programming Interface API

References

1. Pisello, A.L.; Rosso, F.; Castaldo, V.L. The role of building occupants’ education in their resilience to climate-change related events. Energy Build. 2017, 154, 217–231. [CrossRef] 2. Nutt, P.C.; Wilson, D.C. Handbook of Decision Making; John and Wiley & Sons: Hoboken, NJ, USA, 2010. 3. Pomerol, J.-C.; Adam, F. Practical Decision Making–From the Legacy of Herbert Simon to Decision Support Systems. In Actes de la Conference Internationale IFIP TC8/WG8.3; The Monash University Decision Support Systems Laboratory: Prato, Italy, 2004; Volume 3, pp. 647–657. 4. Bonczek, R.H.; Holsapple, C.W.; Whinston, A.B. Foundations of Decision Support Systems; Academic Press: Cambridge, MA, USA, 2014. 5. Turban, E.; Cameron Fisher, J.; Altman, S. Decision support systems in academic administration. J. Educ. Adm. 1988, 26, 97–113. [CrossRef] 6. Rajabifard, A.; Thompson, R.G.; Chen, Y. An intelligent disaster decision support system for increasing the sustainability of transport networks. Nat. Resour. Forum 2015, 39, 83–96. [CrossRef] 7. Ishak, W.H.W.; Ku-Mahamud, K.R.; Morwawi, N.M. Conceptual model of intelligent decision support system based on naturalistic decision theory for reservoir operation during emergency situation. Int. J. Civ. Environ. Eng. 2011, 11, 6–11. 8. Lopes, A.M.G.; Sousa, A.C.M.; Viegas, D.X. Numerical simulation of turbulent flow and fire propagation in complex topography. Numer. Heat Transf. Part A Appl. 1995, 27, 229–253. [CrossRef] 9. Fan, D.; Shi, P. Improvement of Dijkstra’s algorithm and its application in route planning. In Proceedings of the 2010 Seventh International Conference on Fuzzy Systems and Knowledge Discovery, Yantai, China, 10–12 August 2010. 10. Akay, A.E.; Wing, M.G.; Sivrikaya, F. A GIS-based decision support system for determining the shortest and safest route to forest fires: A case study in Mediterranean Region of Turkey. Environ. Monit. Assess 2012, 184, 1391–1407. [CrossRef][PubMed] 11. Correia, A.; Severino, I.; Nunes, I.L. Knowledge management in the development of an intelligent system to support emergency response. In International Conference on Applied Human Factors and Ergonomics; Springer: Berlin/Heidelberg, Germany, 2017. 12. Turban, E.; Sharda, R.; Delen, D. Decision Support and Business Intelligence Systems (Required); Prentice Hall Press: Upper Saddle River, NJ, USA, 2010. 13. Simões-Marques, M.; Figueira, J.R. How Can AI Help Reduce the Burden of Disaster Management Decision-Making? In International Conference on Applied Human Factors and Ergonomics; Springer: Berlin/Heidelberg, Germany, 2018. 14. Alarifi, A.S.N.; Alarifi, N.S.N.; Al-Humidan, S. Earthquakes magnitude predication using artificial neural network in northern Red Sea area. J. King Saud. Univ. Sci. 2012, 24, 301–313. [CrossRef] 15. Goymann, P.; Herrling, D.; Rausch, A. Flood Prediction through Artificial Neural Networks. In Proceedings of the Eleventh International Conference on Adaptive and Self-Adaptive Systems and Applications, At Venice, Italy, 5–9 May 2019. 16. Ardito, L.; Scuotto, V.; Del, G.M. A bibliometric analysis of research on Big Data analytics for business and management. Manag. Decis. 2019, 57, 1993–2009. [CrossRef] 17. Bean, T.H.D. Paul Barth and Randy How Big Data is Different. MIT Sloan Management Review. Available online: https://sloanreview.mit.edu/article/how-big-data-is-different/ (accessed on 12 September 2019). 18. Vecchio, P.D.; Mele, G.; Ndou, V. Creating value from Social Big Data: Implications for Smart Tourism Destinations. Inf. Process. Manag. 2018, 54, 847–860. [CrossRef] 19. Elia, G.; Polimeno, G.; Solazzo, G. A multi-dimension framework for value creation through big data. Ind. Mark. Manag. 2019.[CrossRef] Appl. Sci. 2020, 10, 666 13 of 13

20. Yu, M.; Yang, C.; Li, Y. Big Data in Natural Disaster Management: A Review. Geosciences 2018, 8, 165. [CrossRef] 21. Jeble, S.; Patil, Y. Role of Big Data in Decision Making. Oper. Supply Chain Manag. Int. J. 2018, 11, 36. [CrossRef] 22. Arslan, M.; Roxin, A.; Cruz, C. A Review on Applications of Big Data for Disaster Management. In Proceedings of the 13th International Conference on Signal Image Technology & Internet Based Systems, Jaipur, India, 4–7 December 2017. 23. Simon, H.A. The New Science of Management Decision; Prentice Hall PTR: Upper Saddle River, NJ, USA, 1960. 24. Phillips-Wren, G. Intelligent decision support systems. In Multicriteria Decision Aid and Artiﬁcial Intelligence: Links, Theory and Applications; Wiley-Blackwell: Hoboken, NJ, USA, 2013. 25. Afuah, A.; Tucci, C.L. Value Capture and Crowdsourcing. Acad. Manage. Rev. 2013, 38, 457–460. [CrossRef] 26. Bhatt, P.; Ahmad, A.J.; Roomi, M.A. Social innovation with open source software: User engagement and development challenges in India. Technovation 2016, 52–53, 28–39. [CrossRef] 27. Cappa, F.; Rosso, F.; Hayes, D. Monetary and Social Rewards for Crowdsourcing. Sustainability 2019, 11, 2834. [CrossRef] 28. Lakhani, K.; Wolf, R.G. Why Hackers Do What They Do: Understanding Motivation and Eﬀort in Free/Open Source Software Projects. SSRN Electron. J. 2003.[CrossRef] 29. Cazorla, L.; Alcaraz, C.; Lopez, J. A three-stage analysis of IDS for critical infrastructures. Comput. Secur. 2015, 55, 235–250. [CrossRef] 30. Hayes, D.R.; Cappa, F. Open-source intelligence for risk assessment. Bus. Horiz. 2018, 61, 689–697. [CrossRef] 31. LeCun, Y.; Bottou, L.; Bengio, Y. Gradient-based learning applied to document recognition. Proc. IEEE 1998, 86, 2278–2324. [CrossRef] 32. Goodfellow, I.; Bengio, Y.; Courville, A. Deep Learning; The MIT Press: Cambridge, MA, USA, 2016. 33. LeCun, Y.; Bengio, Y.; Hinton, G. Deep learning. Nature 2015, 521, 436–444. [CrossRef][PubMed] 34. Chollet, F. Keras: The Python Deep Learning Library; Astrophysics Source Code Library, 2018. Available online: https://keras.io/#support (accessed on 18 November 2019).

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).