DMCA
Human-inspired Identification of High-level Concepts using OWA and Linguistic Quantifiers
BibTeX
@MISC{Reformat_human-inspiredidentification,
author = {M. Z. Reformat and R. R. Yager and Z. Li and N. Alajlan and Marek Z. Reformat and Ronald R. Yager and Zhan Li and Naif Alajlan},
title = {Human-inspired Identification of High-level Concepts using OWA and Linguistic Quantifiers},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Abstract: Intelligent agent based system can be used to identify high-level concepts matching sets of keywords provided by users. A new human-inspired approach to concept identification in documents is introduced here. The pro-posed method takes keywords and builds concept structures based on them. These concept structures are represented as hierarchies of concepts (HofC). The ontology is used to enrich HofCs with terms and other concepts (sub-concepts) based on concept definitions, as well as with related concepts. Additionally, the approach uses levels of importance of terms defining the concepts. The levels of importance of terms are continuously updated based on a flow of documents using an Adaptive Assignment of Term Importance (AATI) schema. The lev-els of activation of concepts identified in a document that match these in the HofC are estimated using ordered weighted averaging (OWA) operators with linguistic quantifiers. A simple case study presented in the paper is designed to illustrate the approach.
