Anveshan
A research workspace over the Gujarati archive — it reads the corpus, and never writes to it.
LIVE
RECORD
- ID
- 012
- WORLD
- Knowledge
- TYPE
- Research workspace
- STATUS
- Live
NOTES
The idea
Search finds a passage. Anveshan is for exploring: it pulls people, places and relationships out of the archive, builds a graph you can browse, and never writes back to the archive itself.
Why it exists
To make the corpus explorable and open it to deeper research than one question at a time.
What it does
- An entity page for each person, place or thing, with its relationships, mentions and facts
- Related entities ranked by confidence and how often they appear together
- Merge or ignore to clean up duplicates, with a batch review for near-matches
- A needs-review queue for low-confidence entities, with type correction
- A gazetteer with places shown on interactive maps
- A reader that highlights mentions and links each back to its entity page
DETAILS
- STAGE
- Browsing and curation are built; a collections workspace and a research assistant are planned
- BUILT WITH
- Python, SQLite, NetworkX, Streamlit