Conscience of Blogs: Verifying Contextual Appropriateness of Emotionsbased on Blog Contents

M. Ptaszynski, P. Dybala, W. Shi, R.Rzepka, and K. Araki (Japan)

Keywords

Computational Intelligence, Affect Analysis, IntelligentAgents, Natural Language Processing, Contextual Appropriateness of Emotions

Abstract

This paper presents a method for estimating contextual appropriateness of speaker’s emotions. The method is using an effective analysis system to estimate the speaker’s emotions and a Web mining technique gathering from the Internet associations about emotional common-sense. The baseline of the Web mining technique, using all of the Web, is improved by restricting the query field to the contents of blogs as containing more evaluative information. A conversational agent equipped with this system could choose an appropriate conversational procedure. The proposed method is evaluated using two conversational agents. The use of blog contents improved the method in the aspect of quality as well as time of processing.

Important Links:



Go Back


IASTED
Rotating Call For Paper Image