Human-human communication and knowledge management

Kazuo Sumita

We have been developing knowledge management systems based on natural language processing: a knowledge sharing system analyzing the structure of stored text and supporting NLP based information retrieval, an e-main communication management system generating a summary from e-mail messages, a text mining system extracting causal relations between events and summarizing them, and so on. In order to capture corporate knowledge, it is important to deal with e-mail communication and speech dialogue, because much information and knowledge are exchanged by e-mails and daily face-to-face meetings. This talk presents the knowledge management systems focusing on such human-human communication.