HESML V1R5: Java Library of Ontology-Based Semantic Similarity Measures

Juan José Lastra-Díaz released the Half-Edge Semantic Measures Library (HESML) V1R5 in May 2024. This Java library implements numerous ontology-based semantic similarity measures and Information Content models for WordNet, SNOMED-CT, MeSH, and Gene Ontology. It enables reproducible word and concept similarity experiments without coding via an XML-based input format.

Use Cases

• Benchmarking semantic similarity measures based on the library's collection of implemented models.
• Evaluating pre-trained word embedding models using the library's built-in evaluation capabilities.
• Running reproducible similarity experiments on medical ontologies like SNOMED-CT and MeSH based on the XML-based experiment specification.
• Applying groupwise similarity measures for comparing sets of concepts as mentioned in the description.
• Parsing and working with ontologies in the OBO file format based on the library's support.

Strengths

• Implements most ontology-based semantic similarity measures and Information Content models reported in the literature.
• Supports multiple major ontologies including WordNet, SNOMED-CT, MeSH, and Gene Ontology.
• Provides a novel collection of efficient path-based and groupwise similarity measures.
• Enables reproducible experiments without software coding via an XML-based input format.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The library is distributed under a CC By-NC-SA-4.0 license, restricting commercial use.

Provenance

Source

e-cienciaDatos Harvested Dataverse

Collection Method

Software library release.

Time Range

2024

Freshness

Last updated 2024-05-05 04:08:15.

Geography

null