Academic Commons


Towards More Graceful Interaction: A Survey of Question-Answering Programs

Paris, Cecile L.

One of the aims of Natural Language Processing (NLP) is to facilitate the use of computers by allowing users to interact with systems in natural language. Since such interactions often take the form of question-answering sessions, the process of question-answering is an important part of NLP. In this paper, we are concerned with the progress made towards building question-answering systems which are natural and satisfying to users, allowing them to express themselves freely and answering questions appropriately. Such systems are said to provide graceful interaction. We survey the evolution of question-answering programs, presenting steps and approaches that have been taken towards providing graceful man-machine interaction, and pointing out efficiencies of existing programs and theories. Our emphasis is on the various issues and difficulties encountered in building a user-friendly question-answering program.


More About This Work

Academic Units
Computer Science
Department of Computer Science, Columbia University
Columbia University Computer Science Technical Reports, CUCS-209-85
Published Here
November 7, 2011