Visualization system for analyzing user opinion

Abstract

With the growth of user generated contents (UGC), it is important to know users' opinions about features of emotions quickly. Visualization and clustering are effective methods to observe summary of opinions. In order to decrease users' effort in examining vast amount of UGC, we proposed an interactive interface to visualize dialog data between human and robots. Before visualization, dialog data are analyzed by MeCab and classified by latent Dirichlet allocation (LDA) into several topics. In order to enhance users' perception, this visualization provides a method for coloring sentences with 8 basic colors based on Plutchik's wheel of emotions. This paper explains the developed system with case studies. We expect this visualization assists users to understand user opinions quickly.