Data and method
How the parliamentary archive is built, and the data behind every number on this site.
Downloads
- passages.csv: 781 distinct passages, with every record each was filed under.
- records.csv: 2,047 source documents, with match status and PDF links.
Where the records come from
A scraper searched the Lok Sabha Digital Library for his name and collected 2,047 documents. Each record page links to the original scanned PDF and to the library's own record.
How passages are counted
The scraper extracted 1,930 passages spoken by him. One scanned page often covers several Question Hour items or Rule 377 matters, so the same passage was filed under several titles. Rule used here: Exact text match after collapsing whitespace. That leaves 781 distinct passages (217,153 words): 588 from Question Hour, 189 from debates and 4 from other documents.
Match quality
| Status | Documents |
|---|---|
| speech found | 1,309 |
| whole-day volume: mentions only | 335 |
| mentioned only | 270 |
| tagged but no speaker label in OCR - check manually | 81 |
| search hit, name not in OCR text | 52 |
133 documents are matched by search only, without his name as a speaker in the OCR text. Their record pages carry an “unverified match” banner.
Known gaps and limits
- All text is uncorrected OCR from scanned volumes.
- 1972: 1 passages, although the Lok Sabha sat that year. The scrape may have missed documents.
- 1975: 0 passages, although the Lok Sabha sat that year. The scrape may have missed documents.
- 1980–1984: not a member. Lost the 1980 election (7th Lok Sabha).
- Themes are keyword matches, not hand tags.
Sources
- Parliamentary records: Lok Sabha Digital Library.
- Election results and committee memberships: Lok Sabha member bioprofile; 8th and 9th Lok Sabha results checked against IndiaVotes. 5th and 6th Lok Sabha vote figures are not yet checked.