CSO Classifier 4.0: Software for efficient ontology-based topic annotation in computer science

CSO Classifier 4.0 is an open-source software framework for automatically annotating scholarly documents with research topics from the Computer Science Ontology. The software maps document metadata, including titles, abstracts, and keywords, to ontology concepts through syntactic matching, semantic expansion, and ontology-aware post-processing. This new version 4.0 introduces an updated semantic backbone trained on recent scholarly metadata, soft classification with topic weighting, domain-aware filtering, compatibility with modern Python environments, and improved metadata generation. The package supports scalable, reproducible, and interpretable topic annotation for digital libraries, knowledge graph construction, recommender systems, trend analysis, and science-of-science studies.