Structural analysis of relevance propagation models

XAMENA, EDUARDO; BRIGNOLE, NÉLIDA BEATRIZ; MAGUITMAN, ANA GABRIELA

KNOWLEDGE-BASED SYSTEMS

ELSEVIER SCIENCE BV

Año: 2021 vol. 234 p. 1 - 12

0950-7051

Relevance relations constitute the core of information retrieval. Topical ontologies, such as collaborative webpage classification projects, can provide a basisfor identifying and analyzing such relations. New meaningful relevance relations can be automatically inferred from these ontologies by composing existingones. In this work, several relevance propagation models are analyzed in termsof complex network theory. Structural properties such as Characteristic pathlength, Clustering coefficient and Degree distribution are computed over themodels in order to understand the nature of each underlying network. Thisanalysis raises interesting points about the Small-world and Scale-free structureof some relevance propagation models. Moreover, other connectivity and centrality measures are computed to gain additional insight into the topology ofrelevance. Finally, the analysis is complemented by providing visualizations ofthe k-core decomposition of different relevance propagation models. To illustrate the generalizability of the proposed methodology the analysis is carriedout on an ontology from a different domain. The major theoretical implication of this analysis is the derivation of new instruments to typify semantic networksderived from relevance relations. The results can be exploited in a pragmaticway, as the parameters and properties derived by this analysis can serve as priorknowledge to algorithms for the automatic or semi-automatic construction ofsemantic networks.