Bibliographic dataset from Scopus and reproducible R pipeline for bibliometric analysis of digital technologies, data integration, and decision support in circular economy research (2015–2025)

Description This record contains the bibliographic dataset and reproducible R pipeline used for the study “Digital technologies, data integration, and decision support in circular economy research: A bibliometric analysis (2015–2025)”. The initial Scopus export contained 4,731 records. After content screening, quality control, and formal eligibility checks, the final dataset comprised 3,726 English-language journal articles and reviews published between 2015 and 2025 in 915 sources. Files included_corpus_public.csv contains the public bibliographic dataset, including record identifiers, titles, authors, publication years, sources, DOIs, document types, author and index keywords, and normalized keywords. It does not contain abstracts, cited references, or the licensed raw Scopus export. bibliometric_pipeline.zip contains the R scripts, configuration files, archived screening audit trail, Scopus search query, and generated bibliometric and LDA outputs. The archive has no additional enclosing directory. Screening and quality control All 4,731 records were assessed using their titles, abstracts, and author keywords. Publications were included when circular economy formed part of the research problem or application context and at least one target direction was substantively addressed: digital technologies, data integration, or decision support. The initial assessment included 3,688 records and excluded 1,043. Quality control covered all 1,043 initially excluded records and 614 included records selected for additional checks. Of the 1,657 records reassessed, 62 initially excluded records were included and 19 initially included records were excluded. This produced 3,731 content-screening inclusions. Formal eligibility checks subsequently excluded four records without abstracts and one record associated with an excluded affiliation country, resulting in the final dataset of 3,726 records. OpenAI Codex based on the GPT-5 model was used as a structured review aid during reassessment and quality checking. The model did not modify screening decisions automatically. All final decisions were reviewed and approved by the author. The archived quality-control file documents the reassessment process, while review flags do not imply automatic inclusion. Search query The Scopus search was exported on 11 February 2026 using the following query: TITLE-ABS-KEY ( "circular economy" )AND TITLE-ABS-KEY ("decision support" OR "decision-support" OR"decision making" OR "decision-making" OR"decision support system*" OR "DSS" OR"multi-criteria" OR "MCDA" OR "MCDM" OR"data-driven" OR "data integration" OR"digital technolog*" OR"big data" OR"artificial intelligence" OR "machine learning" OR"internet of things" OR "IoT" OR"blockchain" OR"digital twin" OR"industry 4.0" OR "industry 5.0")AND PUBYEAR > 2014 AND PUBYEAR < 2026AND ( LIMIT-TO ( LANGUAGE, "English" ) )AND ( LIMIT-TO ( DOCTYPE, "ar" ) OR LIMIT-TO ( DOCTYPE, "re" ) )AND EXCLUDE ( AFFILCOUNTRY, "Iran" )AND EXCLUDE ( AFFILCOUNTRY, "Russian Federation" ) Reproducibility The licensed raw Scopus export is not redistributed. Researchers with access to Scopus can reproduce the analysis by exporting the query results as CSV or BibTeX, placing one file named scopus_export* in the archive’s data_raw directory, and running Rscript run_pipeline.R followed by Rscript verify_pipeline.R. The archived pipeline reproduces the final dataset, bibliometric tables and figures, and the nine-topic LDA model.