Visualization of the Meaning of Words on the Web by Extracting Keywords and Emotional Valences

Abstract

The purpose of this study is to show ordinary people the meaning of words on the web. To achieve this purpose, we use cross-search results. The keywords extracted from the sentences constitute nodes. The results are visualized as a tree, the color of which reflects the emotions of the twitter results. To show the effectiveness of our work, we analyze the web traffic of our site, users' comments, a questionnaire about interface design and the results of an experiment comparing our approach with the existing searching way. From these studies, we believe that people will find our work useful as a way to show relationships between words in different linguistic corpora on the web.