Analysis of User Dialogues in the Social Network

With the social networks integration in our lives and becoming an essential part in sharing information, establishing connections and helping others, the problem of analyzing user’s behavior became especially important. Billions of people generate enormous amounts of data that can be used for analysis to train a model which in theory could predict comments or generate them. In this paper we propose analyzing user’s message intent and trying to predict the intent of a following message using deep learning and specific dataset for this problem. Messages can be viewed as a sequences and labelled with intent which converts problem into sequence prediction. We hope the result of this study will contribute to the Social Network Analysis field by allowing other researchers build on what we have accomplished. Keywords: social network analysis, sequence prediction, machine learning, deep learning, recurrent neural networks.