Pub Date : 2013-09-08DOI: 10.1109/SocialCom.2013.32
Sedat Gokalp, M. Temkit, H. Davulcu, I. H. Toroslu
Blogosphere plays an increasingly important role as a forum for public debate. In this paper, given a mixed set of blogs debating a set of political issues from opposing camps, we use signed bipartite graphs for modeling debates, and we propose an algorithm for partitioning both the blogs, and the issues (i.e. topics, leaders, etc.) comprising the debate into binary opposing camps. Simultaneously, our algorithm scales both the blogs and the underlying issues on a univariate scale. Using this scale, a researcher can identify moderate and extreme blogs within each camp, and polarizing vs. unifying issues. Through performance evaluations we show that our proposed algorithm provides an effective solution to the problem, and performs much better than existing baseline algorithms adapted to solve this new problem. In our experiments, we used both real data from political blogosphere and US Congress records, as well as synthetic data which were obtained by varying polarization and degree distribution of the vertices of the graph to show the robustness of our algorithm.
{"title":"Partitioning and Scaling Signed Bipartite Graphs for Polarized Political Blogosphere","authors":"Sedat Gokalp, M. Temkit, H. Davulcu, I. H. Toroslu","doi":"10.1109/SocialCom.2013.32","DOIUrl":"https://doi.org/10.1109/SocialCom.2013.32","url":null,"abstract":"Blogosphere plays an increasingly important role as a forum for public debate. In this paper, given a mixed set of blogs debating a set of political issues from opposing camps, we use signed bipartite graphs for modeling debates, and we propose an algorithm for partitioning both the blogs, and the issues (i.e. topics, leaders, etc.) comprising the debate into binary opposing camps. Simultaneously, our algorithm scales both the blogs and the underlying issues on a univariate scale. Using this scale, a researcher can identify moderate and extreme blogs within each camp, and polarizing vs. unifying issues. Through performance evaluations we show that our proposed algorithm provides an effective solution to the problem, and performs much better than existing baseline algorithms adapted to solve this new problem. In our experiments, we used both real data from political blogosphere and US Congress records, as well as synthetic data which were obtained by varying polarization and degree distribution of the vertices of the graph to show the robustness of our algorithm.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130968686","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-09-08DOI: 10.1109/SocialCom.2013.119
M. Erdmann, Erik Ward, K. Ikeda, Gen Hattori, C. Ono, Y. Takishima
Twitter is a popular medium for sharing opinions on TV programs, and the analysis of TV related tweets is attracting a lot of interest. However, when collecting all tweets containing a given TV program title, we obtain a large number of unrelated tweets, due to the fact that many of the TV program titles are ambiguous. Using supervised learning, TV related tweets can be collected with high accuracy. The goal of our proposed method is to automate the labeling process, in order to eliminate the cost required for data labeling without sacrificing classification accuracy. When creating the training data, we use only tweets of unambiguous TV program titles. In order to decide whether a TV program title is ambiguous, we automatically determine whether it can be used as a common expression or named entity. In two experiments, in which we collected tweets for 32 ambiguous TV program titles, we achieved the same (78.2%) or even higher classification accuracy (79.1%) with automatically labeled training data as with manually labeled data, while effectively eliminating labeling costs.
{"title":"Automatic Labeling of Training Data for Collecting Tweets for Ambiguous TV Program Titles","authors":"M. Erdmann, Erik Ward, K. Ikeda, Gen Hattori, C. Ono, Y. Takishima","doi":"10.1109/SocialCom.2013.119","DOIUrl":"https://doi.org/10.1109/SocialCom.2013.119","url":null,"abstract":"Twitter is a popular medium for sharing opinions on TV programs, and the analysis of TV related tweets is attracting a lot of interest. However, when collecting all tweets containing a given TV program title, we obtain a large number of unrelated tweets, due to the fact that many of the TV program titles are ambiguous. Using supervised learning, TV related tweets can be collected with high accuracy. The goal of our proposed method is to automate the labeling process, in order to eliminate the cost required for data labeling without sacrificing classification accuracy. When creating the training data, we use only tweets of unambiguous TV program titles. In order to decide whether a TV program title is ambiguous, we automatically determine whether it can be used as a common expression or named entity. In two experiments, in which we collected tweets for 32 ambiguous TV program titles, we achieved the same (78.2%) or even higher classification accuracy (79.1%) with automatically labeled training data as with manually labeled data, while effectively eliminating labeling costs.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"111 3S 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131968723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-09-08DOI: 10.1109/SocialCom.2013.108
J. Monti, Mario Monteleone, Maria Pia di Buono, Federica Marano
Extracting relevant information in multilingual context from massive amounts of unstructured, structured and semi-structured data is a challenging task. Various theories have been developed and applied to ease the access to multicultural and multilingual resources. This papers describes a methodology for the development of an ontology-based Cross-Language Information Retrieval (CLIR) application and shows how it is possible to achieve the translation of Natural Language (NL) queries in any language by means of a knowledge-driven approach which allows to semi-automatically map natural language to formal language, simplifying and improving in this way the human-computer interaction and communication. The outlined research activities are based on Lexicon-Grammar (LG), a method devised for natural language formalization, automatic textual analysis and parsing. Thanks to its main characteristics, LG is independent from factors which are critical for other approaches, i.e. interaction type (voice or keyboard-based), length of sentences and propositions, type of vocabulary used and restrictions due to users' idiolects. The feasibility of our knowledge-based methodological framework, which allows mapping both data and metadata, will be tested for CLIR by implementing a domain-specific early prototype system.
{"title":"Natural Language Processing and Big Data - An Ontology-Based Approach for Cross-Lingual Information Retrieval","authors":"J. Monti, Mario Monteleone, Maria Pia di Buono, Federica Marano","doi":"10.1109/SocialCom.2013.108","DOIUrl":"https://doi.org/10.1109/SocialCom.2013.108","url":null,"abstract":"Extracting relevant information in multilingual context from massive amounts of unstructured, structured and semi-structured data is a challenging task. Various theories have been developed and applied to ease the access to multicultural and multilingual resources. This papers describes a methodology for the development of an ontology-based Cross-Language Information Retrieval (CLIR) application and shows how it is possible to achieve the translation of Natural Language (NL) queries in any language by means of a knowledge-driven approach which allows to semi-automatically map natural language to formal language, simplifying and improving in this way the human-computer interaction and communication. The outlined research activities are based on Lexicon-Grammar (LG), a method devised for natural language formalization, automatic textual analysis and parsing. Thanks to its main characteristics, LG is independent from factors which are critical for other approaches, i.e. interaction type (voice or keyboard-based), length of sentences and propositions, type of vocabulary used and restrictions due to users' idiolects. The feasibility of our knowledge-based methodological framework, which allows mapping both data and metadata, will be tested for CLIR by implementing a domain-specific early prototype system.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"84 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132610423","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-09-08DOI: 10.1109/SocialCom.2013.117
Hsin-Tsung Peng, William W. Y. Hsu, Chih-Hung Chen, F. Lai, Jan-Ming Ho
Predicting prices and risk measures of assets and derivatives and rating of financial products have been studied and widely used by financial institutions and individual investors. In contrast to the centralized and oligopoly nature of the existing financial information services, in this paper, we advocate the notion of a Financial Cloud, i.e., an open distributed framework based cloud computing architecture to host modularize financial services such that these modularized financial services may easily be integrated flexibly and dynamically to meet users' needs on demand. This new cloud based architecture of modularized financial services provides several advantages. We may have different types of service providers in the ecosystem on top of the framework. For example, market data resellers may collect and sell long-term historical market data. Statistical analyses of macroeconomic indices, interest rates, and correlation of a set of assets may also be purchased online. Some agencies might be interested in providing services based on rating or pricing values of financial products. Traders may use the statistically estimated parameters to fine-tune their trading algorithm to maximize the profit of their clients. Providers of each service module may focus on effectiveness, performance, robustness, and security of their innovative products. On the other hand, a user pays for exactly what one uses to optimally manage their assets. A user may also acquire services through an online agent who is an expert in assessing the structural model and quality of existing products and thus assembles service modules matching users risk taking behavior. In this paper, we will also present a survey of related existing technologies and a prototype we developed so far.
{"title":"FinancialCloud: Open Cloud Framework of Derivative Pricing","authors":"Hsin-Tsung Peng, William W. Y. Hsu, Chih-Hung Chen, F. Lai, Jan-Ming Ho","doi":"10.1109/SocialCom.2013.117","DOIUrl":"https://doi.org/10.1109/SocialCom.2013.117","url":null,"abstract":"Predicting prices and risk measures of assets and derivatives and rating of financial products have been studied and widely used by financial institutions and individual investors. In contrast to the centralized and oligopoly nature of the existing financial information services, in this paper, we advocate the notion of a Financial Cloud, i.e., an open distributed framework based cloud computing architecture to host modularize financial services such that these modularized financial services may easily be integrated flexibly and dynamically to meet users' needs on demand. This new cloud based architecture of modularized financial services provides several advantages. We may have different types of service providers in the ecosystem on top of the framework. For example, market data resellers may collect and sell long-term historical market data. Statistical analyses of macroeconomic indices, interest rates, and correlation of a set of assets may also be purchased online. Some agencies might be interested in providing services based on rating or pricing values of financial products. Traders may use the statistically estimated parameters to fine-tune their trading algorithm to maximize the profit of their clients. Providers of each service module may focus on effectiveness, performance, robustness, and security of their innovative products. On the other hand, a user pays for exactly what one uses to optimally manage their assets. A user may also acquire services through an online agent who is an expert in assessing the structural model and quality of existing products and thus assembles service modules matching users risk taking behavior. In this paper, we will also present a survey of related existing technologies and a prototype we developed so far.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129544434","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-09-08DOI: 10.1109/SocialCom.2013.149
Matti Mantere
False information spread through online social media and various news outlets can cause significant fluctuations in equity markets around the world. This fluctuation is partially independent of the initial cause of the chain of events that lead to an inaccurate piece of information becoming a widespread rumor. In this paper a method for manipulating stock markets is presented together with a hypothetical case study. The method leverages the way that even unverified information spreads through social and other online media. This is done by intentional dissemination of a made-to-order rumor while simultaneously covertly launching cyber attacks as a catalyst to this process. The intention of this type of activity can is to affect the targeted equity markets for the financial gain of the perpetrators. Through a presentation of a hypothetical case study we argue that the method presented is a viable method for producing illicit gains for criminal groups, and some forms of it might already be in use by some actors.
{"title":"Stock Market Manipulation Using Cyberattacks Together with Misinformation Disseminated through Social Media","authors":"Matti Mantere","doi":"10.1109/SocialCom.2013.149","DOIUrl":"https://doi.org/10.1109/SocialCom.2013.149","url":null,"abstract":"False information spread through online social media and various news outlets can cause significant fluctuations in equity markets around the world. This fluctuation is partially independent of the initial cause of the chain of events that lead to an inaccurate piece of information becoming a widespread rumor. In this paper a method for manipulating stock markets is presented together with a hypothetical case study. The method leverages the way that even unverified information spreads through social and other online media. This is done by intentional dissemination of a made-to-order rumor while simultaneously covertly launching cyber attacks as a catalyst to this process. The intention of this type of activity can is to affect the targeted equity markets for the financial gain of the perpetrators. Through a presentation of a hypothetical case study we argue that the method presented is a viable method for producing illicit gains for criminal groups, and some forms of it might already be in use by some actors.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114307452","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-09-08DOI: 10.1109/SOCIALCOM.2013.35
Bogdan Vasilescu, V. Filkov, Alexander Serebrenik
Stack Overflow is a popular on-line programming question and answer community providing its participants with rapid access to knowledge and expertise of their peers, especially benefitting coders. Despite the popularity of Stack Overflow, its role in the work cycle of open-source developers is yet to be understood: on the one hand, participation in it has the potential to increase the knowledge of individual developers thus improving and speeding up the development process. On the other hand, participation in Stack Overflow may interrupt the regular working rhythm of the developer, hence also possibly slow down the development process. In this paper we investigate the interplay between Stack Overflow activities and the development process, reflected by code changes committed to the largest social coding repository, GitHub. Our study shows that active GitHub committers ask fewer questions and provide more answers than others. Moreover, we observe that active Stack Overflow askers distribute their work in a less uniform way than developers that do not ask questions. Finally, we show that despite the interruptions incurred, the Stack Overflow activity rate correlates with the code changing activity in GitHub.
{"title":"StackOverflow and GitHub: Associations between Software Development and Crowdsourced Knowledge","authors":"Bogdan Vasilescu, V. Filkov, Alexander Serebrenik","doi":"10.1109/SOCIALCOM.2013.35","DOIUrl":"https://doi.org/10.1109/SOCIALCOM.2013.35","url":null,"abstract":"Stack Overflow is a popular on-line programming question and answer community providing its participants with rapid access to knowledge and expertise of their peers, especially benefitting coders. Despite the popularity of Stack Overflow, its role in the work cycle of open-source developers is yet to be understood: on the one hand, participation in it has the potential to increase the knowledge of individual developers thus improving and speeding up the development process. On the other hand, participation in Stack Overflow may interrupt the regular working rhythm of the developer, hence also possibly slow down the development process. In this paper we investigate the interplay between Stack Overflow activities and the development process, reflected by code changes committed to the largest social coding repository, GitHub. Our study shows that active GitHub committers ask fewer questions and provide more answers than others. Moreover, we observe that active Stack Overflow askers distribute their work in a less uniform way than developers that do not ask questions. Finally, we show that despite the interruptions incurred, the Stack Overflow activity rate correlates with the code changing activity in GitHub.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"202 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123224026","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-09-08DOI: 10.1109/SocialCom.2013.84
Amira Bradai, H. Afifi
Host-Based Intrusion Detection Systems (HIDS)have been widely used to detect malicious behaviors of nodes in heterogenous networks. Collaborative intrusion detection can be more secure with a framework using reputation aggregation as an incentive. The problem of incentives and efficiency are well known problems that can be addressed in such collaborative environment. In this paper, we propose to use game theory to improve detection and optimize intrusion detection systems used in collaboration. The main contribution of this paper is that the reputation of HIDS is evaluated before modeling the game between the HIDS and attackers. Our proposal has three phases: the first phase builds reputation evaluation between HIDS and estimates the reputation for each one. In the second phase, a proposed algorithm elects a leader using reputation value to make decisions. In the last phase, using game theory the leader decides to activate or not the HIDS for optimization reasons.
{"title":"Game Theoretic Framework for Reputation-Based Distributed Intrusion Detection","authors":"Amira Bradai, H. Afifi","doi":"10.1109/SocialCom.2013.84","DOIUrl":"https://doi.org/10.1109/SocialCom.2013.84","url":null,"abstract":"Host-Based Intrusion Detection Systems (HIDS)have been widely used to detect malicious behaviors of nodes in heterogenous networks. Collaborative intrusion detection can be more secure with a framework using reputation aggregation as an incentive. The problem of incentives and efficiency are well known problems that can be addressed in such collaborative environment. In this paper, we propose to use game theory to improve detection and optimize intrusion detection systems used in collaboration. The main contribution of this paper is that the reputation of HIDS is evaluated before modeling the game between the HIDS and attackers. Our proposal has three phases: the first phase builds reputation evaluation between HIDS and estimates the reputation for each one. In the second phase, a proposed algorithm elects a leader using reputation value to make decisions. In the last phase, using game theory the leader decides to activate or not the HIDS for optimization reasons.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125093532","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-09-08DOI: 10.1109/SocialCom.2013.120
Caroline Sabty, Martin Memmel, Slim Abdennadher
A growing number of social media services like Flickr and Twitter allows for the association of locations with digital resources such as photos or text messages. This work tackles the problem of exploiting the abundance of such kind of data, by designing and implementing an application that analyzes and visualizes spatial information from different social media services. The application provides two different data retrieval modes, the scenarios and the exploring mode. The scenarios mode accesses information that was already harvested into a local storage, while the exploring mode retrieves information on the fly using the respective APIs. In both modes, information about the returned resources is aggregated on a map and displayed using different visualization means. In addition, the retrieved information is analyzed using three different analysis techniques.
{"title":"GeoEvents - An Interactive Tool to Analyze and Visualize Spatial Information from the Social Web","authors":"Caroline Sabty, Martin Memmel, Slim Abdennadher","doi":"10.1109/SocialCom.2013.120","DOIUrl":"https://doi.org/10.1109/SocialCom.2013.120","url":null,"abstract":"A growing number of social media services like Flickr and Twitter allows for the association of locations with digital resources such as photos or text messages. This work tackles the problem of exploiting the abundance of such kind of data, by designing and implementing an application that analyzes and visualizes spatial information from different social media services. The application provides two different data retrieval modes, the scenarios and the exploring mode. The scenarios mode accesses information that was already harvested into a local storage, while the exploring mode retrieves information on the fly using the respective APIs. In both modes, information about the returned resources is aggregated on a map and displayed using different visualization means. In addition, the retrieved information is analyzed using three different analysis techniques.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"132 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127089694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-09-08DOI: 10.1109/SocialCom.2013.77
Xia Yang, J. Alves-Foss
Security and privacy policies are stated in the context of abstract concepts such as users/roles, objects and actions that relate to a specific level of abstraction in the system design. Refinement of the abstract design down to lower level implementations can result in a disconnect between the implementation and the more abstract security policy. In this paper we introduce the concept of security policy refinement for access control policies that allows us to maintain a tighter coupling between the security policy and its implementation. We use a purpose-based privacy policy as an example to explain the concepts. The resulting refinement technique provides for improved verification and validation that the system, as implemented, satisfies the abstract security policy, and sets the stage for further research in this area.
{"title":"Security Policy Refinement: High-Level Specification to Low-Level Implementation","authors":"Xia Yang, J. Alves-Foss","doi":"10.1109/SocialCom.2013.77","DOIUrl":"https://doi.org/10.1109/SocialCom.2013.77","url":null,"abstract":"Security and privacy policies are stated in the context of abstract concepts such as users/roles, objects and actions that relate to a specific level of abstraction in the system design. Refinement of the abstract design down to lower level implementations can result in a disconnect between the implementation and the more abstract security policy. In this paper we introduce the concept of security policy refinement for access control policies that allows us to maintain a tighter coupling between the security policy and its implementation. We use a purpose-based privacy policy as an example to explain the concepts. The resulting refinement technique provides for improved verification and validation that the system, as implemented, satisfies the abstract security policy, and sets the stage for further research in this area.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"124 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116318524","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2013-09-08DOI: 10.1109/SocialCom.2013.118
Andrey Bogomolov, B. Lepri, F. Pianesi
In this paper we provide the first evidence that daily happiness of individuals can be automatically recognized using an extensive set of indicators obtained from the mobile phone usage data (call log, sms and Bluetooth proximity data) and ``background noise'' indicators coming from the weather factor and personality traits. Our final machine learning model, based on the Random Forest classifier, obtains an accuracy score of 80.81% for a 3-class daily happiness recognition problem. Moreover, we identify and discuss the indicators, which have strong predictive power in the source and the feature spaces, discuss different approaches, machine learning models and provide an insight for future research.
{"title":"Happiness Recognition from Mobile Phone Data","authors":"Andrey Bogomolov, B. Lepri, F. Pianesi","doi":"10.1109/SocialCom.2013.118","DOIUrl":"https://doi.org/10.1109/SocialCom.2013.118","url":null,"abstract":"In this paper we provide the first evidence that daily happiness of individuals can be automatically recognized using an extensive set of indicators obtained from the mobile phone usage data (call log, sms and Bluetooth proximity data) and ``background noise'' indicators coming from the weather factor and personality traits. Our final machine learning model, based on the Random Forest classifier, obtains an accuracy score of 80.81% for a 3-class daily happiness recognition problem. Moreover, we identify and discuss the indicators, which have strong predictive power in the source and the feature spaces, discuss different approaches, machine learning models and provide an insight for future research.","PeriodicalId":129308,"journal":{"name":"2013 International Conference on Social Computing","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114671173","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：