doi.org
The ChinaComx Lianhuanhua Publication Dataset
This is the first release of the ChinaComx Lianhuanhua Publication Dataset, including: Initial Dataset Release: Consolidated lianhuanhua publication database comprising over 38,000 entries. Raw Data: Included original, unparsed OCR extractions from catalogues Ref002 and Ref004 in /data/raw/. Processed Data: Cleaned, structured datasets with disambiguated titles, parsed authors, and standardized publisher locations in /data/processed/. Processing Scripts: Python/Jupyter notebooks for layout analysis, OCR extraction, title splitting, and Transformer-based Named Entity Recognition (NER) author parsing in /scripts/. Repository Infrastructure: Standardized README.md, CITATION.cff, LICENSE.md (CC BY-NC-SA 4.0), and customized GitHub Issue templates.