How ulviyabasova/litreviewkit works
LitReviewKit is a Streamlit workspace for literature discovery and review. Users search scholarly APIs or import Scopus/Web of Science exports; records are normalized, screened against terms, and deduplicated before review. They can manually screen records, inspect project metrics, and export review artifacts. Projects persist in SQLite, while session data is lost on restart. Optional chat uses matching abstracts and an OpenAI key. The repository is a single source file, so component paths share that file; some internal wiring is documented in the README but not visible in the sampled excerpts.
Main entry points: litreviewkit_app.py Live search
Components 20
Discovery and import
- Live search
litreviewkit_app.py - Export import
litreviewkit_app.py - Normalize records
litreviewkit_app.py - Term screening
litreviewkit_app.py - Duplicate merge
litreviewkit_app.py
Review workflow
- Query outcomes
litreviewkit_app.py - Manual screening
litreviewkit_app.py
Project workspace
- Streamlit app
litreviewkit_app.py - Review dashboard
litreviewkit_app.py - Review exports
litreviewkit_app.py - Project management
litreviewkit_app.py - Project database
litreviewkit_app.py - Literature chat
litreviewkit_app.py
External services
- Crossref
- OpenAlex
- Semantic Scholar
- OpenAI API
Other components
- Researcher
- Scopus / WoS files
- Excel / CSV / protocol JSON
Connections 20
- Researcher → Streamlit app: uses
- Streamlit app → Live search: starts search
- Live search → Crossref: queries
- Live search → OpenAlex: queries
- Live search → Semantic Scholar: queries
- Researcher → Export import: selects exports
- Scopus / WoS files → Export import: loads records
- Live search → Normalize records: returns records
- Export import → Normalize records: returns records
- Normalize records → Term screening: screens terms
- Term screening → Duplicate merge: passes matches
- Duplicate merge → Query outcomes: records outcomes
- Query outcomes → Manual screening: presents records
- Manual screening → Review dashboard: updates review
- Manual screening → Review exports: provides review data
- Review exports → Excel / CSV / protocol JSON: writes
- Researcher → Project management: manages project
- Project management → Project database: saves / loads
- Researcher → Literature chat: asks question
- Literature chat → OpenAI API: sends question and abstracts