Please use this identifier to cite or link to this item: http://repositorio.unicamp.br/jspui/handle/REPOSIP/327172
Type: Outros documentos
Title: An Approach To Knowledge Acquisition Based On Verbal Semantics
An approach to knowledge acquisition based on verbal semantics
Author: Vasques, Dildre G.
Zambon, Antonio C.
Baioco, Gisele B.
Martins, Paulo S.
Abstract: This work presents an approach to knowledge acquisition based on semantic classification of verbs as a tool to knowledge management. It allows the extraction of propositions, concepts and non-taxonomic relations from a domain. It also allows a systematic understanding of the knowledge construction process, based on systems thinking and energy flows. We argue that this type of extraction may facilitate the understanding of the structure of cognitive processes and contribute to knowledge extraction in several areas that use representation, storage, transfer and flow of knowledge as a resource. Our experiments show that the proposed approach may extract processual knowledge and represent it in a causal concept map, guaranteeing, as much as possible, the accuracy of the acquired knowledge, by minimizing the distance between the knowledge agent's domain and what the knowledge engineer is capable of extracting.
This work presents an approach to knowledge acquisition based on semantic classification of verbs as a tool to knowledge management. It allows the extraction of propositions, concepts and non-taxonomic relations from a domain. It also allows a systematic
Subject: Gestão do conhecimento
Semântica
Country: Estados Unidos
Editor: Institute of Electrical and Electronics Engineers
Citation: 2016 49th Hawaii International Conference On System Sciences (hicss). Ieee Computer Soc, p. 4144 - 4153, 2016.
Rights: fechado
Fechado
Identifier DOI: 10.1109/HICSS.2016.514
Address: https://ieeexplore.ieee.org/document/7427700
Date Issue: 2016
Appears in Collections:FT - Artigos e Outros Documentos

Files in This Item:
File SizeFormat 
000377358204026.pdf467.35 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.