Semantic relatedness in DBpedia: A comparative and experimental assessment

SemSimp: A Parametric Method for Evaluating the Semantic Similarity of Digital Resources

“…Semantic similarity and the more general notion of semantic relatedness (Formica and Taglino, 2023;Hadj Taieb et al, 2020), is a fundamental research topic in different areas of computer science, for instance in semantic web search (Bollegala et al, 2011;Formica et al, 2010), bioinformatics (Berrhail and Belhadef, 2020;Sharma et al, 2021), crisis management (De Nicola et al, 2019), business processes (De Nicola et al, 2023b), Formal Concept Analysis (Formica, 2019;Wang et al, 2020), Geographic Information Systems (Alizadeh et al, 2021;Formica and Pourabbas, 2009), semantic interoperability (Taglino et al, 2023), etc., however, it is still a challenge. Computing the semantic similarity among textual data (e.g., words, sentences, or documents) is an open research problem in the field of Natural Language Processing (NLP), with several applications ranging from information retrieval and question answering to text summarization and machine translation.…”

Semantic Similarity in a Taxonomy by Refining the Relatedness of Concept Intended Senses

“…In order to compute semantic similarity between concepts, the essential activity consists in evaluating the relatedness [8] between the generic sense and the intended sense of a given concept. With this regard, in this paper, we refine the original proposal of the authors by addressing the ASRMP m relatedness measure proposed in [9], which shows the best correlation with the human judgment concerning other methods defined in the literature [10]. The new experiment leads, for each method addressed in [7], to an average increment of the average correlation with the human judgment of about 0.04.…”

Linked Data Semantic Distance with Global Normalization for evaluating Semantic Similarity in a Taxonomy

“…Note that, in this paper, we address the problem of evaluating both semantic similarity and semantic relatedness by focusing on structured knowledge and, in particular, we rely on ISA taxonomies in the former case and knowledge graphs in the latter case. For this reason, Natural Language Processing (NLP) methods are not addressed, in line with Formica and Taglino (2023a).…”

“…Essentially, with respect to the combIC proposal that leverages triple weights, the LDSDGN γ measure focuses on triple patterns, and aims at verifying the existence of specific configurations of paths in the RDF knowledge graph, Formica and Taglino (2023a).…”