Complex Network Algorithm for Glossary Formation Context-Related Predictive Terms

Abstract

This article describes the process of creating a glossary of terms for a specific domain, which is the initial step in knowledge modeling. In the context of converging trends and interdisciplinary connections in the development of complex systems, particular emphasis is placed on modeling information and communication technologies (ICT) and computer science. To form the glossary of prognostic terms, a comprehensive algorithmic approach was applied, integrating a range of conditions that combine the capabilities of network (graph-based) and semantic approaches. This approach includes automatic graph generation, considering ranking in the evaluation of search results, and context-semantic filtering. As a result, a comprehensive algorithm and software code were developed, allowing the creation of a glossary of contextually related specialized terms and thematic phrases based on the "Wikipedia" network service. These terms were ranked using the average score of two algorithms - PageRank and HITS. The algorithm's operation was visualized using the example of generating a graph from the primary term "Quantum computing". Data were analyzed to justify the objectivity of the proposed term weighting approach and to demonstrate the algorithm's results in expanding the context of prognostic terms within the category of "Computing engineering." A fragment of the structured glossary of ICT is presented as a final demonstration. The results of this research will be used as a foundational knowledge corpus necessary for formulating well-grounded queries when analyzing thematic articles located in bibliographic databases and external network resources.

Keywords

• glossary
• algorithm
• graph
• semantics
• prognostic term
• category
• information and communication technologies