Modeling for Crisis Management Occurred during Computer Mediated Communications

Abstract

Social Network has become a very popular and instrumental procedure in all disciplines for analyzing the web of relationships at an organizational and societal level. This process is enabled by powerful computational skills and data mining techniques. The expertise of this analysis lies in identification of key individuals or groups in societal systems to detect and create network structures and to model its formation and growth. newlineThe information diffused in the human population through randomly established networks behave similar to an epidemic spread. Distribution of people occupying different positions in the network has respective significance. The importance of positions in the network, number of connections and nature of connections, all together determine the metrics of the network. These metrics can further be formalized mathematically to reach to valuable insights. newlineIn this thesis, we have made an effort to contribute effectively in the spread and control of crisis in online communication. First, we have mapped the epidemic model on social networks. Just like epidemic modeling is used to detect and control the spread of an epidemic we have predicted the spread and control the social contagion. Social Contagion is anything that can spread in a population due to interpersonal influence it can be information, norms, behavior, product adoption, common practices, opinions etc. Techniques can be applied to decipher a warning or early detection about adoption of any identifiable practice, norm or behavior in a population. Any information that is harmful to a high fraction of society and spreads through the networks of society is termed to be a crisis in social networks. This has been achieved by addressing three major objectives. newlineFirstly, we have made an effort to map the epidemic model in the prediction in the crisis impact and spread with respect to time. Like any other disease epidemic, the various types of crisis can also be modeled as SIR, SIRS and other epidemic models. We have designed local epidemic mode