DNSS: Dialogue-based news service system in robot environments

Abstract

In this paper, we describe a news service system based on dialogue between human and robot in robot environments. Our goal is to enable users to listen to the news contents more effectively using service robot. Existing news services such as newspaper, TV news program, and Internet news depend on their media. People who want to acquire news contents must be skilled to deal with the media in the case of Internet news service. Also, such media as newspaper and TV also have time and space constraints. In order to solve these problems, we propose a dialogue-based news service which provides news contents through spoken language dialogue. For dialogue news service, a news knowledge model based on ontology was designed. The knowledge model is centered on not articles but news events. People are able to reorganize the contents of known events, and talk to others with brief sentences. Likewise, this system analyzes articles about events, reorganizes news contents and provides the news through spoken language dialogue. For more effective service, a personalized user model was suggested, which reflects not only topic preferences but also interaction patterns. The developed system called Dialogue-based News Service System (DNSS) provides personalized news contents using the user model and the news knowledge model.