INTERACTIVE SEMANTIC ANALYSIS OF TECHNICAL TEXTS

Sentence syntax is the basis for organizing semantic relations in TANK A, a project that aims to acquire knowledge from technical text. Other hallmarks include an absence of preceded domain-specific knowledge; s ignificant use of public-domain generic linguistic information sources ; involvement of the user as a judge and source of expertise; and lear ning from the meaning representations produced during processing. Thes e elements shape the realization of the TANKA project: implementing a trainable text processing system to propose correct semantic interpret ations to the user. A three-level model of sentence semantics, includi ng a comprehensive Case system, provides the framework for TANKA's rep resentations. Text is first processed by the DIPETT parser, which can handle a wide variety of unedited sentences. The semantic analysis mod ule HAIKU then semi-automatically extracts semantic patterns from the parse trees and composes them into domain knowledge representations. H AIKU's dictionaries and main algorithm are described with the aid of e xamples and traces of user interaction. Encouraging experimental resul ts are described and evaluated.