Automatic Concept Classifiication of Text From Electronic Meetings

Chen, Hsinchun and Hsu, P. and Orwig, Richard E. and Hoopes, L. and Nunamaker Jr., Jay F. (1994) Automatic Concept Classifiication of Text From Electronic Meetings. Communications of the ACM 37(10):pp. 56-73.

Abstract

In this research we adopted an artificial intelligence (AI) approach to designing an automatic concept classification tool for electronic brainstorming output. The role of AI techniques such as machine learning and neural networks computing in groupware development can be significant. Through extensive content analysis, concept space generation, and neural network-based concept classification, our system can generate a tentative list of the important ideas and topics represented in meeting comments. Participants then can examine the system’s suggested list and the underlying comments. They can also revise or augment the list to produce their final consensus list. Allowing the system to act as an “intelligent” aide for idea organization can alleviate some of the burdens of convergent tasks.