CNIRS Chinese News IR System

The project is now a public portfolio-ready Chinese news retrieval system: the live demo exposes the interactive Flask app and API stats, while the portfolio page packages the cover, screenshots, WebM walkthrough, technical explanation, and interview flow.

• Supports BM25, TF-IDF, Boolean, Hybrid, LM, BIM, WAND, MaxScore, Fuzzy, and CSoundex retrieval modes.
• Facet browsing works without a query, so users can directly browse all news matching selected metadata.
• Each result exposes snippets, highlights, component scores, field boosts, and Why this result explanations.
• Document detail combines summary, KWIC, keywords, taxonomy metadata, related news, and source facets.
• Model Comparison, Evaluation, Ranking Diagnostics, Feedback Analytics, and Analysis Graph pages make the IR pipeline visually demonstrable.

• Consolidating research-style IR modules into a stable, testable, demo-ready service API.
• Handling Chinese tokenization, normalization, metadata cleaning, taxonomy mapping, and facet quality.
• Keeping heavy optional models such as CKIP/BERT out of low-resource startup paths through graceful fallback.
• Presenting a dynamic Flask system from a static GitHub Pages portfolio using screenshots, video, and a dothost live demo link.

• Portfolio reviewers and interviewers
• Technical readers who need a quick view of purpose, stack, and maturity

• Detected technical signals: Python, Flask, JavaScript, Bootstrap, Jieba, BM25, TF-IDF, Boolean Retrieval, Language Model Retrieval, SQLite, Chart.js, Playwright, pytest
• README evidence exists and can support a fuller reviewed case study
• A public GitHub repository is linked for source traceability

This case study is generated from the portfolio catalog pipeline using README, Git metadata, package/build configuration, and media signals. The final architecture narrative still needs source-level review. Current detected technology signals include: Python, Flask, JavaScript, Bootstrap, Jieba, BM25, TF-IDF, Boolean Retrieval, Language Model Retrieval, SQLite, Chart.js, Playwright, pytest.

A public data-flow narrative is not fully reviewed yet. If the project includes data processing, AI pipelines, or backend APIs, the next pass should document input, processing, storage, and UI/output flow end to end.

information-retrieval/
  README.md              # project documentation, when available
  source files           # implementation reviewed by local audit
  package/build config   # detected capability signals

This project does not expose a verified runnable web command yet. Review the README/source tree and add exact install, run, test, and build commands before interview use.
No verified build command was detected. Treat the current portfolio page as a case-study placeholder until build steps are reviewed.

• Complete the production-quality README, screenshots, and demo recording
• Add architecture diagrams, data-flow notes, and key technical decisions
• Verify build/test status and update the portfolio release report

• State the current maturity and demonstrable scope first
• Focus on verified stack, source structure, and completed behavior
• Do not claim unverified deployment, video, or test coverage as finished

• Expand qrels and human relevance labels for a stronger evaluation benchmark.
• Grow the feedback dataset into a real learning-to-rank experiment.
• Expose semantic retrieval as an optional production profile rather than a default startup dependency.