SNS data Visualization for analyzing spatial-temporal distribution of social anxiety

Abstract

In this paper, we describe SNS (Social Networking Service, especially Twitter) data visualization for analyzing spatial-temporal distribution of social anxiety. We prepare train data collected from Twitter by using Open API(twitter4j), which represent whether the person who post Tweet, posting message in Twitter, is anxious or not. From these data, dictionary explaining frequency of words is constructed by using KOMORAN which is Korean morphological analysis library. And we design classifier based on Naive Bayes method and estimate degree of anxiety of Tweet which include spatial-temporal information. We visualize these estimations as the form of web application, which are represented as a map and word cloud. As the spatial-temporal data are visualized in this way, we can analyze public opinion about a variety of social events.