CosMovis: Analyzing semantic network of sentiment words in movie reviews

In this paper, we present the new method to easily recognize intricate network and cluster by connecting Multidimensional Scaling(MDS) Map and Social Network Graph, and comprehend the feature of each node using Heatmap Visualization. For the sake of this method, this paper used netizen´s movie review data. Also, the process is as follow : 1) We calculated the frequency of sentiment word from each movie review. 2) We designed Heatmap Visualization which easily apprehends main emotion of netizens which appear in each movie review. 3) We made location of each node reflect the frequency of sentiment word by designing Sentiment-Movie Network made up of combining MDS Map and Socal Network Graph, as well as we embodied Network Graph to make same nodes form cluster. 4) By granting meaning in accordance with characteristic of clustering, we applied asterism graphic to facilitate cognitive interpretation. Our demonstration is available to http://idlab.ajou.ac.kr/cosmovis/.