Data for: Agentic GeoAI — A Bibliometric Mapping and Typology of Conversational Spatial Intelligence

This dataset contains the bibliometric records underlying "Agentic GeoAI: A Bibliometric Mapping and Typology of Conversational Spatial Intelligence." It consists of two corpora retrieved from the OpenAlex API in July 2026: 528 records for classical GeoAI (spatially explicit machine learning for prediction/classification) and 1,509 records for LLM–geospatial literature (natural-language, tool-using, agentic systems applied to geospatial problems), covering publication years 2015–2026. Three records with confirmed corrupted OpenAlex metadata (mismatched titles and citation counts on reused DOIs) were identified and excluded prior to analysis. Each corpus file (classical_geoai.json, agentic_geoai.json) contains per-work metadata: OpenAlex ID, DOI, title, publication year, venue, citation count, OpenAlex concept tags, author count, author-affiliation countries, work type, and open-access status. The accompanying CSV files are the derived summary tables reported in the paper (publication counts by year, top venues, top concepts, most-cited works, country distribution by corpus), and citation_details.json provides full author/venue details for the specific papers cited by name in the paper's typology section. Source metadata originates from OpenAlex (openalex.org), released under CC0.