2000 Volume 24 Issue 2 Pages 143-152
The authors conduct an effectiveness research of e-mail discussion review support system with summary extraction method. The support system we have developed can automatically extract summary sentences from the colloquial e-mail messages using reference relationship of e-mails they have discussed. Japanese natural language processing technology is applied in the proposed method. In order to evaluate the effectiveness of the system, we conducted some experiment using questionnaire method and protocol analysis. We compared the two system ; the system with and without summary sentences in the table of e-mail content. As a result, following fact-findings were obtained. The system with summary sentences could promote reading strategy such as utilization of table of contents and comprehension of e-mail message structures. On the other hand, the system without summary sentence makes the reader give attention to the detail information such as names. From above described result, we concluded that the system with summary sentence is effective to understandings of relationship of e-mail messages.
