Integration of Narrative Processing, Data Fusion, and Database Updating Techniques in an Automated System

摘要

Narrative data processing and database updating are examined from the viewpoint of interaction with other processes in an automated data fusion system. Methods are outlined for using production rules (if-then rules) in the processing of the narrative data and for integrating such a process with other rule-based data-fusion processes. Database updating methods which employ a 'history file' are processed. An experimental rule-based program compares event reports for redundancies, combines reports of the same event and retires old data into a history file in a form useful for event reconstruction and evaluation. A number of related areas of investigation are suggested for future research. (Author)