knowledge-paper-extractor

You can extract metadata and references from a scientific PDF by providing the file and an output directory. The tool identifies title, authors, abstract, and cited references, generating structured metadata.json, references.json, and summary.md files.

Parsing citations for a systematic review is best done using refextract to collect references and Crossref to resolve them into CSL-JSON. This creates structured references.json and references.csl.json files for knowledge management.

Yes, you can extract the abstract and authors from a PDF. The tool uses pdfminer.six for heuristic analysis of the first page and pypdf to pull embedded metadata, detecting the title, authors, and abstract automatically.

Yes, the tool resolves extracted PDF references into CSL-JSON format. It uses Crossref to resolve the references collected by refextract, outputting a references.csl.json file alongside standard metadata and summary outputs.

When extracting bibliographic data from PDFs, the generated output formats include metadata.json, references.json, references.csl.json, and summary.md. These structured files support literature reviews and knowledge management workflows.

Heuristic analysis detects the title and abstract from the first page of a PDF using pdfminer.six to evaluate text layout and content. It pulls structural information to separate the title, authors, and abstract accurately.