OpenAlex Academic Metadata Dataset for Hybrid Knowledge-Graph-Enhanced Retrieval-Augmented Generation Framework

This dataset contains a filtered subset of approximately 5,000 scholarly records extracted from the OpenAlex academic database. The dataset includes scholarly metadata fields such as titles, abstracts, authors, institutional affiliations, publication venues, topics, keywords, citation information, and referenced works. The dataset was prepared for the development and evaluation of a Hybrid Knowledge-Graph-Enhanced Retrieval-Augmented Generation (KG-RAG) framework for semantic information modeling in academic digital libraries. The dataset supports semantic embedding generation, knowledge graph construction, vector indexing, hybrid retrieval, and response generation experiments.