A NATURAL-LANGUAGE PROCESSING BASED GROUP DECISION-SUPPORT SYSTEM

Group Decision Support Systems (GDSSs) are currently used primarily as surface-level discussion tools. That is, current GDSSs do not allow group members to easily access information in deeper levels, such as the data base, model base, and application programs. This paper describes ways in which GDSSs can be improved by using a natural language interface to allow group members to communicate with deeper-level information systems using human languages. The system consists of database, model base, application programs and natural language interface system. The system is designed to both route questions to appropriate subsystems and translate these questions into the computer language controlling these subsystems. Finally, experimental results demonstrate the feasibility of the technique.