Replication Data for: Revealing the shifts in mineral governance and sustainability policies: A multi-method text mining approach in Türkiye (1963–2028)

This repository contains the comprehensive raw textual dataset, custom Python source code, and supplementary materials supporting the manuscript "Revealing the shifts in mineral governance and sustainability policies: A multi-method text mining approach in Türkiye (1963–2028)."

The dataset comprises 65 years (1963–2028) of national planning documents, including 12 National Development Plans and 14 Special Expert Commission Reports (SECRs) on mining, coal, and metal minerals. The provided Python script allows for the full reproduction of the multi-method NLP pipeline, including regex-based preprocessing, TF-IDF signature keyword extraction, Latent Dirichlet Allocation (LDA) topic modeling, semantic co-occurrence network matrix generation, and lexicon-based risk sentiment analysis. These materials, along with the domain-specific stop-word lists, are shared to promote strict transparency and computational reproducibility in mineral governance and policy research.