Corpus Index for Lexical vs Semantic Retrieval Experiment in Biomedical Literature

This dataset provides the corpus index used in the controlled biomedical literature retrieval experiment reported in the associated article on lexical and semantic retrieval. The file contains article-level identifiers and metadata to support corpus reconstruction and independent verification of the study sample. Specifically, it includes: PMCID PMID Article title The indexed corpus was used to compare Bag-of-Words (BoW), TF-IDF, and embedding-based semantic retrieval methods within a fixed experimental setting. The purpose of releasing this index is to improve transparency and reproducibility while respecting copyright and licensing restrictions on the underlying full-text content. Important: The repository does not include article full text or abstracts. Researchers seeking to reproduce the analysis must obtain access to the corresponding PubMed-indexed records and, where required, full-text articles in accordance with publisher and database licensing requirements. This dataset is intended as a reconstruction aid for the controlled corpus rather than as a redistribution of the source literature.