RUS  ENG
Full version
JOURNALS // Informatika i Ee Primeneniya [Informatics and its Applications] // Archive

Inform. Primen., 2021 Volume 15, Issue 3, Pages 83–90 (Mi ia747)

This article is cited in 5 papers

Forms representing new knowledge discovered in texts

I. M. Zatsman

Federal Research Center “Computer Science and Control” of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation

Abstract: The model of goal-oriented discovery of new knowledge by a team of experts in texts and the forms of its representation in linguistic typologies and term dictionaries of medical knowledge bases are considered. The usage of the model is illustrated by two examples: the discovery of new knowledge about the meanings of German modal verbs in parallel texts and about the meanings of terms in medical documents. The process of discovering new knowledge is based on linguistic annotation of texts performed by experts. The main goal of annotation is to enrich typologies with headings or, accordingly, augment knowledge bases with terms that meet the specified criterion of novelty, are agreed on within the team of experts, and represent the discovered knowledge. The model includes the coordination of experts' understanding of both the discovered knowledge and the forms of its representation. In the example of the term dictionary of a medical knowledge base, three forms are used: a new term, a changed definition of an existing term without increasing the number of its meanings, and an enlarged definition of an existing term with its new meanings.

Keywords: discovery of knowledge in texts, ITO model, concept, heading, term, contextual meaning, German modal verbs, typology, medical knowledge base.

Received: 14.07.2021

DOI: 10.14357/19922264210311



© Steklov Math. Inst. of RAS, 2024