Using Social Networks Data for Behavior and Sentiment Analysis

With the advent of social networks, a huge amount of information is generated and stored every day. The social networks therefore represent a potentially infinite source of user data, usable both for scientific and commercial applications. Specifically, they store a lot of data regarding the single individual and behavior, as well as information related to individuals and their relationship with other individuals, i. e. a sort of collective behavior. The combination of behavior and sentiment analysis tools with methodologies of affective computing could allow the extraction of useful data that convey information for different applications, such as detection of depression state in psychology, political events, stock marketing fluctuations. The paper surveys some data extraction tools for social networks and emerging computing trends such as behavior analysis, sentiment analysis and affective computing. Then the paper proposes a first architecture for behavior analysis integrating those tools.