Affiliations: [a] KDDI Corporation, Iidabashi 3-10-10, Chiyoda, Tokyo, 102-8460, Japan | [b] Graduate School of Buisiness Sciences, University of Tsukuba, Otsuka 3-29-1, Bunkyo, Tokyo 112-0012, Japan
Abstract: The prospect of wide spreading the Internet in the world is enabling new ways of solving our several problems with many Question and Answering web sites, Q&A sites. It is difficult to say that smooth communication is carried out in these sites, because a lot of speakers may speak inconsistently in the same site at the same time. Therefore, we proposed to extract the egocentrism from language expressions in words at some Japanese Q&A sites and tried to express the mutual understanding gap in these words based on the egocentrism. Specifically, we defined the intensity for properties of the egocentrism with the method that presumed the egocentrism from language expressions in our previous study. This method expresses mutual understanding gap in the text data of the sites that have dialogue form by using the intensity and calculating the score of the egocentrism in each speech. This paper shows the shape of graphs with these scores expresses the characteristics of the gap. It also evaluates this method with text data in actual Q&A sites and confirms its effectiveness, describing the precision is 98.6%. It also analyzes features of the mutual understanding gap at the site. It shows that the score of the mutual understanding gap is evenly distributed and most communications are attempted by the few speeches.
