cnki-advanced-search

Automating CNKI advanced search uses Playwright to construct staged keyword-group queries, filter CSSCI journals, sort by citations, and export a deduplicated categorized Excel summary file.

The three-stage literature retrieval strategy builds an onion-style query plan: main joint search, independent supplementary background search, and final merged deduplication to collect both core and background literature.

When multi-keyword CNKI advanced searches return zero or overly small outputs, the workflow applies automatic keyword-group downgrading to a minimum of two groups and triggers mandatory reflection and adjustment.

Yes, Playwright and openpyxl are required dependencies. Playwright drives browser automation for CNKI navigation and CAPTCHA/login detection, while openpyxl generates the final deduplicated Excel output file.

Yes, the workflow executes independent supplement retrieval for background and theory literature alongside direct intersection literature, merging both batches into a unified categorized spreadsheet.

CNKI automated search failures during CSSCI filtering often stem from CAPTCHA or login prompts blocking Playwright browser automation, which the workflow detects and handles to maintain the export flow.