Ontologically Structured Methods for Evaluating Semantic Textual Similarity in Security Applications
Atul Kumar Gupta, Rohit Saxena, Vishal Nagar, Satyasundara Mahapatra · 2025
In computer engineering and cognitive science, evaluating semantic textual similarity (STS) among phrases, paragraphs, and documents is crucial. Additionally, it has numerous uses in a variety of industries, including geomatics and bioinformatics. In this chapter, we analyzed the concept of semantic recognition of patterns and the utilization of the tool that is basically concerned about security domain. In this study, we give a review on several linguistic similarity techniques, as well as information on the existence of various STS-related tools and apps. For several tasks in natural language processing (NLP), including abstractive summarization, semantic features, short response grading, pattern recognition, and extracting, STS is a critical part. They categorize the conceptual similarity metrics into three main groups: knowledge-based similarity, corpus-based similarity, and string-based similarity. The techniques connected to the WordNet ontology are highlighted more. Structural approaches are crucial for deciphering the original meaning of ambiguous terms because they are exceedingly challenging for computers to analyze. We also suggest a novel outlook to gauging sentence semantic relatedness. This suggested approach combines the information into a language model while utilizing the benefits of taxonomy methodologies.