Proceedings of the IEEE

Natural language interactions with artificial experts

Tim Finin, Aravind K. Joshi, and Bonnie Webber

July 1, 1986

The aim of this paper is to justify why Natural Language (NL) interaction, of a very rich functionality, is critical to the effective use of Expert Systems and to describe what is needed and what has been done to support such interaction. Interactive functions discussed here include defining terms, paraphrasing, correcting misconceptions, avoiding misconceptions, and modifying questions. Natural Language interfaces to database systems are already proving their worth. They allow users to get at the database facts they want, without the need to become system wizards. In this paper, we are primarily concerned with NL interfaces for systems that do more than identify and retrieve facts. Such systems, often called knowledge-based systems, expert systems, or advisory systems, are expected to provide analyses and/or advice to users faced with real problems. Our main goals in this paper are to justify why NL interaction, of a very rich functionality, is critical to the effective use of these systems and to demonstrate what is needed to support it. Even if one wanted to employ a formal language or menu and pointer-based system for this role, it would have to have many of the features of NL that allow the kinds of interactive functions that a system must support.

BibTeX OWL Tweet Scholar

Tags: ai.nlp, artificial intelligence, computational linguistics, expert system, natural language processing

Type: Article

Publisher: IEEE

Organization: IEEE

Volume: 74

Number: 7

Note: DOI: 10.1109/PROC.1986.13572

Downloads: 2953 downloads

Google Scholar Citations: 30 citations