Thematic Subjects and SDG Tagging – Human vs AI Indexing Dataset (Unicamp IR)

This dataset contains records from the Institutional Repository of the State University of Campinas (Unicamp) used to compare human indexing with AI-based classification. Provided in a single Excel file (.xlsx), it includes two sheets: subjects_dataset: Subject terms manually assigned by librarians from the Unicamp Library System for 40 scientific documents. sdgs_dataset: Sustainable Development Goals (SDGs) automatically assigned to the same documents using a generative AI tool based on the Google Gemini model. The dataset enables comparative analyses of conceptual consistency, term relevance, and thematic classification accuracy between human and AI-generated metadata, with applications in academic library workflows and metadata quality assessment.