090°Open data
Open datasets, fully documented — searchable here, and readable by any LLM.
1 datasets
House of Commons of Canada (Hansard)
Per-MP, per-month debate sentiment and topic engagement from the official House of Commons Hansard (45th Parliament, 1st session; keyless XML). Each row aggregates one MP's English floor interventions in a calendar month: participation counts, a documented lexicon-based sentiment read (fixed parliamentary word lists, 3-token negation window; French translations excluded via FloorLanguage gating), a 0-100 sentiment_score (50 = neutral) ranked within the month, and topic-engagement concentration (Herfindahl over debate topics) with the top topic named. Speakers are entity-resolved by the House's stable DbId person id; titled affiliations (ministers, the PM, parliamentary secretaries) resolve to the same person record. Units: counts, words, shares, scores. Caveats: months with no sittings (e.g. summer recess) are absent, never zero-filled; MP-months under 3 interventions / 100 words are dropped; French-original passages appear in the -E edition in official English translation. Primary key: (month, mp_dbid). Cadence: monthly; as_of is the latest sitting date, so identical input produces an identical content hash. Hansard text is House of Commons parliamentary material (Crown copyright; commercial_use = unclear) — only aggregate statistics are published, never speech text.
Once today's free allowance is used up, AI features can run on your own provider account.
Kept in this browser tab only (cleared when you close it) and sent with each AI request. Our servers use it for that request and never store or log it.