Natural Language Processing in the FAQ Finder System: Results and Prospects

This paper describes some recent results regarding the employment of natural language processing techniques in the FAQ FINDER system. FAQ FINDER is a natural language question-answering system that uses files of frequently-asked questions as its knowledge base. Unlike AI questionanswering systems that focus on the generation of new answers, FAQ FINDER retrieves existing ones found in frequently-asked question files. FAQ FINDER uses a combination of statistical and natural language techniques to match user questions against known question/answer pairs from FAQ files. We strove in our experiments to identify the contribution of these techniques to the overall success of the system. One unexpected result was that our parsing technique was not contributing as much to the system’s performance as we expected. We discuss some of the reasons why this may have been the case, and describe further natural language processing research designed to address the system’s current needs.