3 commits

Author	SHA1	Message	Date
Matt	620f99c762	Add business_intent_poi_boost reranker signal ... When a query contains no road-type keywords (st, blvd, ave, etc.), boost amenity/shop/tourism/leisure/office/craft results (+3.0) and penalize highway/route results (-4.0). This fixes searches like "starbucks twin falls" where a named service road outranked the actual business POI due to Photon position tiebreaking. Also fixes: - Intent classifier now recognizes full state names ("idaho" not just "ID") for LOCALITY classification - Locality-type Photon results now populate _city from name field so they participate in locality_fuzz scoring - Trace logging expanded to all candidates with osm_key/value Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-04-21 19:39:37 +00:00
Matt	d460f0e202	Fix type classifier: POI check takes precedence over street_address ... Businesses with housenumbers (e.g. M&W Markets at 130 US-30) were classified as street_address because the housenumber check fired before the osm_key check. Reorder so osm_key in amenity/shop/tourism/leisure/office is evaluated first, ensuring businesses get type=poi regardless of whether they have a street address. Also adds office to the POI key set. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-04-21 19:08:04 +00:00
Matt	87b230dcba	feat(navi): structured geocode with usaddress parsing and reranker ... Add lib/geocode.py — multi-source retrieval pipeline: - usaddress CRF parsing with intent classification - Netsyms structured lookup (uses raw street abbreviations) - Photon /structured + /api freetext retrieval - Weighted 10-signal reranker (housenumber, street fuzz, locality, source authority, etc.) - match_code annotations + address book proximity labeling - Trace log at /tmp/geocode_rerank_trace.log nav_tools.py now delegates geocode() to the new module. Tests updated: US address queries correctly return Netsyms results. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-04-20 16:29:47 +00:00