Knowledge Acquisition Using Linguistic-Based Knowledge Analysis

Most knowledge-based system development efforts include acquiring knowledge from one or more sources. Difficulties associated with this knowledge acquisition task are readily acknowledged by most researchers. While a variety of knowledge acquisition methods have been reported, little has been done to organize those different methods and to suggest how to apply them within a conceptual framework. The linguistic-based knowledge analysis appreach described here offers a conceptual approach wherein knowledge in a subject area is analyzed as (1) lexical, (2) syntactic, and (3) semantic knowledge. Analysis of these three separate components of knowledge creates a domain “language” that completely describes the knowledge of a particular subject area in logical detail. This “language” serves as a knowledge model that can be readily translated into a software implementation. The resulting knowledgebased system is able to “converse” in the language of its domain about particular types of problems. Many previously reported knowledge acquisition techniques are effective, to varying degrees, for acquiring lexical, syntactic, and semantic knowledge. Vol. 12, Nos. 1-3, 1998 Knowledge Acquisition Using Linguistic-Based Knowledge Analysis Daniel L. Schmoldt USDA Forest Service Brooks Forest Products Center Virginia Tech Blacksburg, Virginia 24061-0503