Skip to main content

Address Data Sources for NER Training

Catalog of public-domain and openly-licensed data sources containing facility names + addresses, intended as raw material for trainingtrainingThe process of adjusting a model's parameters so its predictions match labeled examples, by repeatedly measuring error and nudging the weights to reduce it. Distinct from inference, when the trained model is run on new input. a tokentokenOne word or subword in the tokenized input. For the neural classifier, tokens come from SentencePiece (subword units); for the rule classifiers, tokens are whitespace- and punctuation-separated words.-classification (BIO-tagging) postal address parser. Companion to federal/aggregator sources (OpenAddressesOpenAddresses (OA). A global open aggregation of address points collected from many official sources. A primary source of component-supervised training data outside proprietary registries., OSMOpenStreetMap (OSM). A community-curated global map database (ODbL-licensed) with addr:* tagged features and place hierarchies. A secondary corpus source and a source of street names., NADNAD (National Address Database). A US Department of Transportation dataset of structured address points, added to the training corpus as a major source of real US addresses., TIGERTIGERThe US Census Topologically Integrated Geographic Encoding and Referencing database. Used as a corpus source for street-segment data., NPI, IRS BMF, etc.) — this list focuses on stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-level sources that add component-level variety, residential coveragecoverageThe fraction of a population or region for which a data source has real, non-placeholder entries — e.g. 47% rooftop coverage on Texas addresses. Distinct from accuracy on the rows that are present., and venuevenueA named, non-address place — a business, building, park, or stadium. Mailwoman's free-text point-of-interest component, added as a Tier 2 fine label.-name-paired records the federal sources lack.

Compiled with web-search verification of canonical agency URLs. Some per-stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. entries in the Regulated Facilities section are based on known agency patterns rather than direct URL verification — those are noted inline.

Why state sources matter for an address NER model​

  1. Component supervision is automatic. StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. directories are typically already field-structured (separate columns for streetstreetThe named linear feature along which house numbers are ordered. Decomposes into a name plus street affixes; one of the Tier 2 fine labels., citylocalityThe city / town / settlement component of an address: a populated place sitting between region and neighbourhood in the hierarchy., postcodepostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one., etc.). BIO tags derive from which column each tokentokenOne word or subword in the tokenized input. For the neural classifier, tokens come from SentencePiece (subword units); for the rule classifiers, tokens are whitespace- and punctuation-separated words. came from — no manual labeling required.
  2. Format diversity per-stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.. 50 statesregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. + DC means 50+ schema flavors, abbreviation conventions, and transcription styles — useful raw material for trainingtrainingThe process of adjusting a model's parameters so its predictions match labeled examples, by repeatedly measuring error and nudging the weights to reduce it. Distinct from inference, when the trained model is run on new input. format robustness.
  3. VenuevenueA named, non-address place — a business, building, park, or stadium. Mailwoman's free-text point-of-interest component, added as a Tier 2 fine label.+address pairs are everywhere. Every facility/professional/establishment listing pairs an organization name with a structured address — natural B-venue / I-venue trainingtrainingThe process of adjusting a model's parameters so its predictions match labeled examples, by repeatedly measuring error and nudging the weights to reduce it. Distinct from inference, when the trained model is run on new input. spansspanA contiguous range of characters or tokens in the input string, tagged with an address component type (street, locality, postcode, etc.). Parsed addresses are represented as collections of spans, possibly nested in a tree..
  4. Residential coveragecoverageThe fraction of a population or region for which a data source has real, non-placeholder entries — e.g. 47% rooftop coverage on Texas addresses. Distinct from accuracy on the rows that are present.. Real estate, contractor, cosmetology, notary, and license-lookup tools surface home/office addresses that federal facility datasets miss.

Formats to watch for beyond CSV​

Many stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. agencies distribute address data in geospatial formats. The DBF attribute table inside a shapefile is essentially a CSV with geometry attached — extract with ogr2ogr -f CSV and the records are trainingtrainingThe process of adjusting a model's parameters so its predictions match labeled examples, by repeatedly measuring error and nudging the weights to reduce it. Distinct from inference, when the trained model is run on new input.-ready. Common formats:

  • Shapefile (.shp/.dbf/.shx/.prj) — most common stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. GIS distribution
  • GeoJSON — increasingly standard on ArcGIS Hub portals
  • GeoPackage (.gpkg) — modern SQLite-based, multi-layerlayerOne transformer block — attention plus a feed-forward network, with normalization and residual connections — applied to every position. Stacking layers lets the model build up richer representations; Mailwoman's encoder has 6.
  • File Geodatabase (.gdb) — ESRI native, common from stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. GIS offices
  • KML / KMZ — Google Earth, sometimes used for facility points
  • FlatGeobuf — emerging streaming format (used by OpenAddressesOpenAddresses (OA). A global open aggregation of address points collected from many official sources. A primary source of component-supervised training data outside proprietary registries.)

The highest-value geospatial layerslayerOne transformer block — attention plus a feed-forward network, with normalization and residual connections — applied to every position. Stacking layers lets the model build up richer representations; Mailwoman's encoder has 6. for address NER:

  • Site address pointssitus dataA dataset of exact address-point coordinates (rooftop-level). Mailwoman's geocoder uses a national situs layer (124.9M US points built from state address-point sources) as the highest-precision tier of the geocode cascade. / E911E911 (Enhanced 911). County- and state-level emergency-dispatch address-point databases with full component breakdown. A lineage source for the situs and address-point layers used in geocoding. / NG911NG911 (Next Generation 911). The modern 911 data standard. States and counties publish NG911 address-point layers that Mailwoman can ingest as address data. address pointssitus dataA dataset of exact address-point coordinates (rooftop-level). Mailwoman's geocoder uses a national situs layer (124.9M US points built from state address-point sources) as the highest-precision tier of the geocode cascade. — every addressable location as a point with full component breakdown (housenumber, prefix, streetstreetThe named linear feature along which house numbers are ordered. Decomposes into a name plus street affixes; one of the Tier 2 fine labels., suffix, unitunitA subdivision of a building — apartment, suite, floor — that refines a street address. Mailwoman's unit component; a designator plus identifier forms a subpremise., citylocalityThe city / town / settlement component of an address: a populated place sitting between region and neighbourhood in the hierarchy., ZIP, parcelparcelA property polygon or record carrying a situs (site) address and often a separate owner mailing address. County GIS parcel aggregations are a training source for address-point variety and situs-vs-owner divergence. ID). Published by counties and increasingly by statesregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. via ArcGIS Hub. The strongest single layerlayerOne transformer block — attention plus a feed-forward network, with normalization and residual connections — applied to every position. Stacking layers lets the model build up richer representations; Mailwoman's encoder has 6. type for component-supervised trainingtrainingThe process of adjusting a model's parameters so its predictions match labeled examples, by repeatedly measuring error and nudging the weights to reduce it. Distinct from inference, when the trained model is run on new input..
  • ParcelparcelA property polygon or record carrying a situs (site) address and often a separate owner mailing address. County GIS parcel aggregations are a training source for address-point variety and situs-vs-owner divergence. polygons with situssitusThe physical site address of a property, as opposed to the owner's mailing address. Parcel records often carry both; the divergence is a real-world data-quality challenge. + owner addresses — two addresses per record (property situssitusThe physical site address of a property, as opposed to the owner's mailing address. Parcel records often carry both; the divergence is a real-world data-quality challenge. + owner mailing), often diverging structurally (P.O. boxesPO boxA numbered mailbox at a post office used as a delivery address instead of a physical street location. Mailwoman tags it as the po_box component; structurally the same family as a subpremise. for owners, streetstreetThe named linear feature along which house numbers are ordered. Decomposes into a name plus street affixes; one of the Tier 2 fine labels. addresses for situssitusThe physical site address of a property, as opposed to the owner's mailing address. Parcel records often carry both; the divergence is a real-world data-quality challenge.). StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-aggregated in MA, NC, MN, UT, MT, VT, RI, and a handful of others.
  • Road centerlines with address ranges — stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-current analogue to TIGERTIGERThe US Census Topologically Integrated Geographic Encoding and Referencing database. Used as a corpus source for street-segment data.; each segmentsegmentA punctuation-bounded chunk of the normalized input — the comma-separated parts of 'Portland, OR' — used to give downstream stages structural context. has from/to house numbershouse numberThe numeric or alphanumeric identifier of a building on a street. Mailwoman's house_number component; its position relative to the street name flips between locales. per side.
  • Critical infrastructure / facilities layerslayerOne transformer block — attention plus a feed-forward network, with normalization and residual connections — applied to every position. Stacking layers lets the model build up richer representations; Mailwoman's encoder has 6. — stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. equivalents of HIFLD covering hospitals, schools, fire/police, jails, courts, public buildings. Each is a B-venue + address pair.
  • EPA/stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-DEP regulated sites — Superfund, brownfields, RCRA, NPDES, leaking USTs as shapefile/GeoJSON.
  • NCES, NREL AFDC, FCC ASR/ULS — federal but distributed as shapefile in addition to CSV.

Most stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. ArcGIS Hubs publish data under CC-BY or equivalent open terms; check per-dataset.

Licensing reality​

StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. records aren't "public domain" in the IP sense, but factual data (name + address) isn't copyrightable under Feist v. Rural Telephone. A few caveats:

  • Washington OSPI school directory: stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. law forbids commercial use of directory data.
  • OSMOpenStreetMap (OSM). A community-curated global map database (ODbL-licensed) with addr:* tagged features and place hierarchies. A secondary corpus source and a source of street names. (ODbL): share-alike obligations may extend to modelsneural classifierThe machine learning model at the core of Mailwoman's parser — a transformer encoder (~30M parameters) trained from scratch to do BIO token classification over addresses. It learns the 'grammar' of address formats; the gazetteer supplies the 'atlas.' trained on it — unsettled, but a risk if the corpuscorpusThe BIO-labeled training data used to train Mailwoman's neural classifier. Assembled from real sources (OpenAddresses, National Address Database) and synthetic shards (boundary stress, order variants, negative space). Managed by @mailwoman/corpus. is republished.
  • CC-BY stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. ArcGIS Hub data: attribution required if redistributing the corpuscorpusThe BIO-labeled training data used to train Mailwoman's neural classifier. Assembled from real sources (OpenAddresses, National Address Database) and synthetic shards (boundary stress, order variants, negative space). Managed by @mailwoman/corpus..
  • G-NAFG-NAF (Geocoded National Address File). Australia's authoritative open address register (CC-BY-licensed), used as a training source for Australian addresses., CC-BY-4.0: clean for AU coveragecoverageThe fraction of a population or region for which a data source has real, non-placeholder entries — e.g. 47% rooftop coverage on Texas addresses. Distinct from accuracy on the rows that are present..

For maximum licensing safety on published model weightsmodel weightsThe learned parameters of the neural classifier, shipped as ONNX files in the @mailwoman/neural-weights-* packages. Weights are locale-specific bundles that include the model, tokenizer, and a model-card.json metadata file., restrict to US-federal sources (NCES, NADNAD (National Address Database). A US Department of Transportation dataset of structured address points, added to the training corpus as a major source of real US addresses., TIGERTIGERThe US Census Topologically Integrated Geographic Encoding and Referencing database. Used as a corpus source for street-segment data., NPPES, IRS BMF) + stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. sources confirmed as public records with no use restrictions. Lose coveragecoverageThe fraction of a population or region for which a data source has real, non-placeholder entries — e.g. 47% rooftop coverage on Texas addresses. Distinct from accuracy on the rows that are present., gain unambiguous redistribution rights.

Gazetteer / resolver coordinate sources (companion layer)​

Everything below this section is corpuscorpusThe BIO-labeled training data used to train Mailwoman's neural classifier. Assembled from real sources (OpenAddresses, National Address Database) and synthetic shards (boundary stress, order variants, negative space). Managed by @mailwoman/corpus. material — address strings and name/address pairs to train the BIO parser. This section is the other half: the gazetteergazetteerA geographical index that maps place names and postcodes to real-world coordinates. Mailwoman uses a custom-built Who's On First (WOF) SQLite database as its gazetteer — the 'atlas' half of the grammar/atlas architecture. the resolverresolverThe component that converts parsed address components (locality, region, postcode) into coordinates by looking them up in the gazetteer. The resolver ranks candidates by name match, population, and proximity, and returns the best-matching place with its centroid or polygon. and the postcode anchoranchor inferenceA technique where structured knowledge (postcode locations, gazetteer place names) is injected into the model as soft input features — not as deterministic overrides. The model still decides the final labels, but the anchor signal biases it toward correct admin tags. consume — the datasets that turn a parsed spanspanA contiguous range of characters or tokens in the input string, tagged with an address component type (street, locality, postcode, etc.). Parsed addresses are represented as collections of spans, possibly nested in a tree. into a place + coordinate. Different shopping list, and it's where WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations.'s gaps bite (see resolver-wof-sqlite/POSTCODE-ANCHOR.md and the 2026-06-03 postcodepostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one.-anchor postmortem).

WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. is an excellent admin gazetteergazetteerA geographical index that maps place names and postcodes to real-world coordinates. Mailwoman uses a custom-built Who's On First (WOF) SQLite database as its gazetteer — the 'atlas' half of the grammar/atlas architecture. (the in-DB hierarchy + ancestors table is our Elasticsearch-free differentiator), and it's CC0, which is half of why it came naturally. But it was never complete for every tokentokenOne word or subword in the tokenized input. For the neural classifier, tokens come from SentencePiece (subword units); for the rule classifiers, tokens are whitespace- and punctuation-separated words. type: no street node (neighbourhood → address directly), thin and uneven postcodepostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one. geometry outside a few countries (US/NL carry own coords; DE ~66% placeable via ancestor-borrow; ES/IT orphan-heavy and effectively unplaceable), and limited POIpoint of interest (POI). A named place that is not strictly an address — landmark, transit stop, venue, amenity, or franchise. Mailwoman tags these as venue and resolves them through the gazetteer./venuevenueA named, non-address place — a business, building, park, or stadium. Mailwoman's free-text point-of-interest component, added as a Tier 2 fine label. coveragecoverageThe fraction of a population or region for which a data source has real, non-placeholder entries — e.g. 47% rooftop coverage on Texas addresses. Distinct from accuracy on the rows that are present.. The canonical fix is the rest of the PeliasPeliasAn open-source geocoder, Mailwoman's spiritual predecessor. four-layerlayerOne transformer block — attention plus a feed-forward network, with normalization and residual connections — applied to every position. Stacking layers lets the model build up richer representations; Mailwoman's encoder has 6. stack — WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. (admin) + OpenAddressesOpenAddresses (OA). A global open aggregation of address points collected from many official sources. A primary source of component-supervised training data outside proprietary registries. (addresses/postcodespostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one.) + OpenStreetMapOpenStreetMap (OSM). A community-curated global map database (ODbL-licensed) with addr:* tagged features and place hierarchies. A secondary corpus source and a source of street names. (streetsstreetThe named linear feature along which house numbers are ordered. Decomposes into a name plus street affixes; one of the Tier 2 fine labels./POIspoint of interest (POI). A named place that is not strictly an address — landmark, transit stop, venue, amenity, or franchise. Mailwoman tags these as venue and resolves them through the gazetteer.) + GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. (places/postcodespostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one.) — supplemented per gap.

By gap layer​

  • PostcodepostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one. → centroid. GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. postal (download.geonames.org/export/zip, CC-BY 4.0, ~80+ countries) is a ready-made postcodepostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one. → place + admin + lat/lon table, the cheapest fill for the DE/ES/IT gap with no point-cloud math (coarser than OA in places, often localitylocalityThe city / town / settlement component of an address: a populated place sitting between region and neighbourhood in the hierarchy.-level). OpenAddressesOpenAddresses (OA). A global open aggregation of address points collected from many official sources. A primary source of component-supervised training data outside proprietary registries. point aggregation (median centroid per postcodepostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one.) is higher-fidelity where we have the points. National authorities are the gold standard but fragmented: UK Code-Point Open (OS, OGL), France BANBAN (Base Adresse Nationale). France's authoritative open national address register — the highest-quality training source for French addresses, with full component structure., NL via PDOK.
  • StreetstreetThe named linear feature along which house numbers are ordered. Decomposes into a name plus street affixes; one of the Tier 2 fine labels. / rooftoprooftopGeocoding precision at the building or parcel level — coordinates within a few metres — the highest tier of the geocode cascade. Sourced from address-point and situs data. coords. OpenAddressesOpenAddresses (OA). A global open aggregation of address points collected from many official sources. A primary source of component-supervised training data outside proprietary registries. (primary), OSMOpenStreetMap (OSM). A community-curated global map database (ODbL-licensed) with addr:* tagged features and place hierarchies. A secondary corpus source and a source of street names. addr:* on buildings (great in DE/NL, patchy elsewhere), national registries — France BANBAN (Base Adresse Nationale). France's authoritative open national address register — the highest-quality training source for French addresses, with full component structure., Australia G-NAFG-NAF (Geocoded National Address File). Australia's authoritative open address register (CC-BY-licensed), used as a training source for Australian addresses. (CC-BY; backlog #31), Sweden Lantmäteriet (national address register; access gated on Geotorget approval — shelved pending that external clock, fills the SE gap).
  • POIspoint of interest (POI). A named place that is not strictly an address — landmark, transit stop, venue, amenity, or franchise. Mailwoman tags these as venue and resolves them through the gazetteer. / landmarks / salience (the generalized-anchor and exotic-POIpoint of interest (POI). A named place that is not strictly an address — landmark, transit stop, venue, amenity, or franchise. Mailwoman tags these as venue and resolves them through the gazetteer. layerlayerOne transformer block — attention plus a feed-forward network, with normalization and residual connections — applied to every position. Stacking layers lets the model build up richer representations; Mailwoman's encoder has 6.). OSMOpenStreetMap (OSM). A community-curated global map database (ODbL-licensed) with addr:* tagged features and place hierarchies. A secondary corpus source and a source of street names. is the richest open POIpoint of interest (POI). A named place that is not strictly an address — landmark, transit stop, venue, amenity, or franchise. Mailwoman tags these as venue and resolves them through the gazetteer. source. Wikidata / Wikipedia is the right home for the data-driven salience signal the anchor guardrail demands (notability + coordinates + multilingual names), and we already pull wikimedia-importance.csv in build-importance, so the path has precedent.
  • Admin (the anchor). Keep WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations.. Alternatives if it is ever reconsidered: GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. admin, Overture Divisions, geoBoundaries.

Probe: GeoNames vs the WOF postcode gap (2026-06-03)​

Measured GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. postal against the gap in the WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. shardshardA partial output file of the corpus build, written in Parquet format. The training pipeline streams shards row by row. (the postcodespostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one. WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. carries as membership but cannot place). GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. carries coordinates for 100% of its DE/ES/IT records:

localelocaleThe combination of language and country an address comes from. en-US and fr-FR are the locales Mailwoman ships weights for.WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. placementGeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. closesresult
IT0% (orphans + wrong links)4,447 / 4,936 (90%)and it fixes the bad links — Milan 20121 → 45.46, 9.19 (correct), where WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. pointed at a Liguria village
DE66% (ancestor-borrow)4,819 / 10,061 unplaced (~48%)would lift DE to ~82%
ESnot built (orphan-heavy)~full (11,150 codes, all with coords)the clean way to add ES at all
NL100% (own PC6 coords)no help — GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. is PC4-level (4,086), coarser than WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations.'s PC6keep WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations.

So GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. postal is the cheap, CC-BY fill for ES/IT (and roughly half of DE's remaining gap), and it corrects WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations.'s Italian mis-links as a side effect. Integration stays clean under the discipline above: match on the postcodepostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one. string, keep the WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. id, write the GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. lat/lon as the centroid. (Overture was not probed — for postcodespostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one. it folds in OpenAddressesOpenAddresses (OA). A global open aggregation of address points collected from many official sources. A primary source of component-supervised training data outside proprietary registries., which we already use, so its differential value is the POIpoint of interest (POI). A named place that is not strictly an address — landmark, transit stop, venue, amenity, or franchise. Mailwoman tags these as venue and resolves them through the gazetteer. layerlayerOne transformer block — attention plus a feed-forward network, with normalization and residual connections — applied to every position. Stacking layers lets the model build up richer representations; Mailwoman's encoder has 6.; that probe wants a DuckDB + S3-parquetParquetThe open columnar file format the corpus is written and streamed in. The training pipeline reads shards row-by-row from Parquet. setup and is the right next step for the generalized-anchor work, not for postcodespostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one..)

The modern entrant: Overture Maps​

The Meta/Microsoft/AWS/TomTom consortium (2023+, post-dates the corpuscorpusThe BIO-labeled training data used to train Mailwoman's neural classifier. Assembled from real sources (OpenAddresses, National Address Database) and synthetic shards (boundary stress, order variants, negative space). Managed by @mailwoman/corpus. catalog below). Its Places theme is a large open POIpoint of interest (POI). A named place that is not strictly an address — landmark, transit stop, venue, amenity, or franchise. Mailwoman tags these as venue and resolves them through the gazetteer. dataset and its Addresses theme folds in OpenAddressesOpenAddresses (OA). A global open aggregation of address points collected from many official sources. A primary source of component-supervised training data outside proprietary registries., so it covers two of our three gaps in one source, with stable GERS entity IDs. Distributed as cloud parquetParquetThe open columnar file format the corpus is written and streamed in. The training pipeline reads shards row-by-row from Parquet. (DuckDB/S3). The one to probe before committing to per-source OSMOpenStreetMap (OSM). A community-curated global map database (ODbL-licensed) with addr:* tagged features and place hierarchies. A secondary corpus source and a source of street names. imports.

Licensing gradient​

For an AGPL product, license terms decide more than coveragecoverageThe fraction of a population or region for which a data source has real, non-placeholder entries — e.g. 47% rooftop coverage on Texas addresses. Distinct from accuracy on the rows that are present. does:

SourceLicenseFriction for a shipped DB
WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations.CC0none (public domain)
OpenAddressesOpenAddresses (OA). A global open aggregation of address points collected from many official sources. A primary source of component-supervised training data outside proprietary registries.per-source, mostly openper-source attribution tracking (we already do this)
GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps.CC-BY 4.0attribution
OvertureCDLA-Permissive 2.0attribution
OpenStreetMapOpenStreetMap (OSM). A community-curated global map database (ODbL-licensed) with addr:* tagged features and place hierarchies. A secondary corpus source and a source of street names.ODbLattribution + share-alike on derived databases — the spicy one; can pull copyleft on

So the gradientgradientThe direction and rate at which the loss would change if each parameter were nudged. Training follows the gradient downhill to reduce error. Huge gradients are tamed by gradient clipping. argues for GeoNamesGeoNamesA free global gazetteer combining administrative, postal, and POI data across 200+ countries. Supplements Who's On First for postcode centroids and places where WOF has gaps. + Overture + OpenAddressesOpenAddresses (OA). A global open aggregation of address points collected from many official sources. A primary source of component-supervised training data outside proprietary registries. as the supplements, and treating OSMOpenStreetMap (OSM). A community-curated global map database (ODbL-licensed) with addr:* tagged features and place hierarchies. A secondary corpus source and a source of street names. carefully (derived signals, not the shipped DB).

Integrity discipline​

Attach supplemental data as attributes on WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations.-keyed entities (a centroid here, a popularity score there), not as imported foreign-id entities. WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. stays the anchor and the evalevalRunning the model against a held-out golden dataset and computing per-component F1, exact-match, calibration, and resolved-coordinate error. keys; the coordinates and the long tail come from elsewhere. Mixing in Overture's GERS ids (or any parallel id space) as primary keys is what would break the WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations.-id-keyed resolverresolverThe component that converts parsed address components (locality, region, postcode) into coordinates by looking them up in the gazetteer. The resolver ranks candidates by name match, population, and proximity, and returns the best-matching place with its centroid or polygon. evalsevalRunning the model against a held-out golden dataset and computing per-component F1, exact-match, calibration, and resolved-coordinate error., which is the reason behind the "extend the custom WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. build, never a prebuilt dump" rule. OA centroid aggregation passes this test because the postcodepostcodeThe country-specific postal code (US ZIP, French code postal, etc.). Mailwoman handles postcode parsing entirely by rule classifier — a regex problem, not an ML one. keeps its WOFWOF (Who's On First). An open-source gazetteer of places maintained by Mapzen/whosonfirst. Mailwoman builds a custom SQLite database from WOF GeoJSON repos, extended with postcode data, importance scores, and coincident-role relations. id; only its coordinate comes from OA.

Education​

Federal aggregators​

  • NCES Common Core of Data (CCD) — Public School & District Universe — https://nces.ed.gov/ccd/ccddata.asp — Bulk CSV/SAS downloads of every public school + district in the US with address. Default CCD home: https://nces.ed.gov/ccd/ — public-use, no restriction.
  • NCES CCD Public School Name & Address File — https://nces.ed.gov/ccd/psadd.asp — Direct CSV address files, grouped by stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.. Public domain.
  • NCES Private School Universe Survey (PSS) — https://nces.ed.gov/surveys/pss/ — Biennial directory of private schools (not bundled into CCD).

State school directories​

Libraries​

Federal aggregator​

State public library directories​

Licensed professionals​

StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-level public databases of licensed professionals (with practice/office addresses) suitable as trainingtrainingThe process of adjusting a model's parameters so its predictions match labeled examples, by repeatedly measuring error and nudging the weights to reduce it. Distinct from inference, when the trained model is run on new input. data for a postal address parsingaddress parsingThe process of decomposing a free-text postal address string into structured components — house number, street name, locality, region, postcode, and country — so a geocoder can resolve them to coordinates. NER modelneural classifierThe machine learning model at the core of Mailwoman's parser — a transformer encoder (~30M parameters) trained from scratch to do BIO token classification over addresses. It learns the 'grammar' of address formats; the gazetteer supplies the 'atlas.'. All URLs verified via web search. Most stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. license-lookup tools are query-only (search-by-name/number); the NPPES NPI Registry and several stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. portals offer bulk CSV exports.

Federal / multi-state​

  • NPPES NPI Registry (CMS, all US healthcare providers w/ practice addresses) — https://npiregistry.cms.hhs.gov/ — query-only UI + REST API
  • NPPES NPI Downloadable Files (full monthly + weekly increments, >4 GB CSV) — https://download.cms.gov/nppes/NPI_Files.html — bulk download (best single source)
  • NMLS Consumer Access (mortgage loan originators, multi-stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality., CSBS) — https://www.nmlsconsumeraccess.org/ — query-only
  • DocInfo / FSMB (physician profiles, sourced from stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. boards) — https://www.docinfo.org/ — query-only
  • FSMB Physician Data Center (PDC) (data files available for purchase) — https://www.fsmb.org/PDC/pdc-data-files/ — bulk (paid)
  • Nursys (NCSBN, multi-stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. RN/LPN verification) — https://www.nursys.com/ — query-only
  • AIM DocFinder (multi-stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. physician metadata aggregator) — http://docfinder.docboard.org/ — query-only
  • ARELLO (multi-stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. real estate licensee directory) — https://www.arello.com/ — query-only

State unified license portals​

(Some statesregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. route many professions through a single search; per-profession entries below reference these where applicable.)

Per state​

Alabama​

Alaska​

Arizona​

Arkansas​

California​

Colorado​

Connecticut​

Delaware​

  • All four via DELPROS — https://delpros.delaware.gov/OH_VerifyLicense — query-only
  • Medical: Delaware Board of Medical Licensure & Discipline (via DELPROS)
  • Nursing: Delaware Board of Nursing (via DELPROS)
  • Real estate: Delaware Real Estate Commission (via DELPROS)
  • Contractor: Local-only at general-contractor level; specialty trades (electrical, plumbing) licensed via DELPROS

District of Columbia​

Florida​

Georgia​

Hawaii​

Idaho​

Illinois​

  • All four via Illinois IDFPR License Lookup — https://online-dfpr.micropact.com/lookup/licenselookup.aspx — query-only
  • Medical: IL Medical Board (via IDFPR)
  • Nursing: IL Board of Nursing (via IDFPR)
  • Real estate: IL Division of Real Estate (via IDFPR)
  • Contractor: Illinois does not have a statewide general contractor license — local-only at GC level; roofing & plumbing licensed via IDFPR

Indiana​

  • All via Indiana PLA — https://secure.in.gov/apps/pla/search and https://mylicense.in.gov/everification/ — query-only
  • Medical: Indiana Medical Licensing Board (via PLA)
  • Nursing: Indiana State Board of Nursing (via PLA)
  • Real estate: Indiana Real Estate Commission (via PLA)
  • Contractor: Local-only — no stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. general contractor license (Plumbing Commission licenses plumbers via PLA)

Iowa​

Kansas​

  • Medical: Kansas Board of Healing Arts — https://www.kansas.gov/ssrv-ksbhada/search.html — query-only
  • Nursing: Kansas State Board of Nursing — https://portal.egov.ksbn.ks.gov/verification/ — query-only
  • Real estate: Kansas Real Estate Commission — https://www.krec.ks.gov/resources/search — query-only
  • Contractor: Local-only — no stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. contractor license (stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. regulates only roofers via AG; specialty/trade licensure varies)

Kentucky​

Louisiana​

Maine​

Maryland​

Massachusetts​

Michigan​

  • All four via MiPLUS / LARA — https://val.apps.lara.state.mi.us/ — query-only
  • Medical: MI Board of Medicine (via MiPLUS)
  • Nursing: MI Board of Nursing (via MiPLUS)
  • Real estate: MI Board of Real Estate Brokers & Salespersons (via MiPLUS)
  • Contractor: MI Residential Builders & M&A Contractors (via MiPLUS)

Minnesota​

Mississippi​

Missouri​

  • All via MO DPR Licensee Search — https://pr.mo.gov/licensee-search.asp — query-only; bulk "Downloadable Listings" updated nightly at https://pr.mo.gov/
  • Medical: Missouri Board of Registration for the Healing Arts (via DPR)
  • Nursing: Missouri State Board of Nursing (via DPR)
  • Real estate: Missouri Real Estate Commission (via DPR)
  • Contractor: Local-only — no stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. general contractor license

Montana​

Nebraska​

Nevada​

New Hampshire​

  • All four via NH OPLC License Lookup — https://www.oplc.nh.gov/license-lookup — query-only
  • Medical: NH Board of Medicine (via OPLC)
  • Nursing: NH Board of Nursing (via OPLC)
  • Real estate: NH Real Estate Commission (via OPLC)
  • Contractor: Local-only — no stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. general contractor license (NH licenses some trades, e.g. electricians, via OPLC)

New Jersey​

New Mexico​

New York​

North Carolina​

North Dakota​

Ohio​

Oklahoma​

Oregon​

Pennsylvania​

  • All medical/nursing/real estate via PALS — https://www.pals.pa.gov/ — query-only
  • Medical: PA State Board of Medicine (via PALS)
  • Nursing: PA State Board of Nursing (via PALS)
  • Real estate: PA Real Estate Commission (via PALS)
  • Contractor: PA Attorney General Home Improvement Contractor Registration — https://hicsearch.attorneygeneral.gov/ — query-only

Rhode Island​

South Carolina​

South Dakota​

Tennessee​

Texas​

Utah​

  • All four via Utah DOPL License Lookup Verification — https://secure.utah.gov/llv/search/index.html — query-only
  • Medical: Utah Physician Licensing Board (via DOPL)
  • Nursing: Utah Board of Nursing (via DOPL)
  • Real estate: Utah Division of Real Estate (via DOPL) — separate division but uses DOPL lookup
  • Contractor: Utah Contractors (via DOPL)

Vermont​

Virginia​

Washington​

West Virginia​

Wisconsin​

  • All four via Wisconsin DSPS License Search — https://licensesearch.wi.gov/ — query-only
  • Medical: WI Medical Examining Board (via DSPS)
  • Nursing: WI Board of Nursing (via DSPS)
  • Real estate: WI Real Estate Examining Board (via DSPS)
  • Contractor: WI Dwelling Contractor / trades (via DSPS)

Wyoming​

  • Medical: Wyoming Board of Medicine — https://wyomedboard.wyo.gov/consumers/license-lookup — query-only
  • Nursing: Wyoming State Board of Nursing — https://wybn.boardsofnursing.org/licenselookup — query-only
  • Real estate: Wyoming Real Estate Commission — https://realestate.wyo.gov/ (public license search) — query-only
  • Contractor: Local-only — no stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. general contractor license (electricians licensed at stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. level via State Fire Marshal)

Secretary of State, business, notary, charity, lobbyist​

Notes on scope and quirks​

  • Business entity searches: every stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. offers free name/agent query; only a minority offer bulk downloads (FL, OR partial via open data, LA via paid API, HI by-record paid). DE does not publish principal addresses. NV charges for some access; TX SOSDirect charges $1/search.
  • Notary registries: most statesregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. publish searchable directories. NH, TN, WY publish only PDF lists or none. GA notaries are commissioned by county clerks but indexed centrally by GSCCCA.
  • Charity registries: typically AG, but SoS in CO, GA, MD, MS, NC, ND, OK, PA, SC, TN, WA, WV; DBR in RI; DCP in CT; DLCP in DC; VDACS in VA; DPFR in ME; DFI in WI; FDACS in FL; DCP in UT (transitioning to DCCC). DE, ID, MT, SD, WY have no stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-level charity registration (only fundraisers in some). AZ repealed charity registration in 2013 (veterans + fundraisers still register).
  • Lobbyist registries: SoS for most; Ethics Commission for AL, GA, HI, KS, MO, OK, SC, TN, TX, WI, WV; Legislature for IA, NE, NV; PIC for DE; APOC for AK; DLS Ethics Council for VA; PDC for WA; OGEC for OR; ELEC for NJ; BEGA for DC; JLEC/OLIG for OH; KLEC for KY; LREC/ILRC for IN; ME Ethics; MA SoC; NH SoS; MD SEC.

Per state​

Alabama​

Alaska​

Arizona​

Arkansas​

California​

  • Business entity search: California Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (bizfile Online) — https://bizfileonline.sos.ca.gov/search/business — query-only; includes principal + agent addresses; free PDF copies
  • Notary registry: California Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://www.sos.ca.gov/notary/notary-public-listing — query-only (active commissions)
  • Charity registry: California Attorney General Registry of Charities and Fundraisers — https://rct.doj.ca.gov/Verification/Web/Search.aspx?facility=Y — query-only
  • Lobbyist registry: California Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (CAL-ACCESS) — https://cal-access.sos.ca.gov/Lobbying/ — query-only; bulk daily data downloads available

Colorado​

Connecticut​

Delaware​

District of Columbia​

Florida​

Georgia​

Hawaii​

Idaho​

  • Business entity search: Idaho Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (SOSBiz) — https://sosbiz.idaho.gov/search/business/ — query-only
  • Notary registry: Idaho Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (SOSBiz) — https://sosbiz.idaho.gov/search/notary — query-only
  • Charity registry: ID has NO general charity registration requirement. No public registry. Investigate manually. Telephone solicitors register with AG: https://www.ag.idaho.gov/
  • Lobbyist registry: Idaho Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (Sunshine system) — https://sunshine.voteidaho.gov/lobbyists — query-only

Illinois​

  • Business entity search: Illinois Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://apps.ilsos.gov/businessentitysearch/ — query-only
  • Notary registry: Illinois Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://apps.ilsos.gov/notarysearch/ — query-only
  • Charity registry: Illinois Attorney General (Charitable Trust Bureau) — https://charitable.illinoisattorneygeneral.gov/ — query-only (new portal launched Aug 2025)
  • Lobbyist registry: Illinois Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://apps.ilsos.gov/lobbyistsearch/index.jsp — query-only; bulk download via Illinois Data Portal

Indiana​

Iowa​

Kansas​

Kentucky​

Louisiana​

Maine​

Maryland​

Massachusetts​

Michigan​

Minnesota​

Mississippi​

Missouri​

Montana​

  • Business entity search: Montana Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://biz.sosmt.gov/search — query-only
  • Notary registry: Montana Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://sosmt.gov/notary/find-a-notary/ — query-only (opt-in for public listing)
  • Charity registry: MT has NO stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-level charity registration. No public registry. Investigate manually
  • Lobbyist registry: Montana Commissioner of Political Practices (COPP) — https://politicalpractices.mt.gov/home/Legislative-Session-and-Lobbying/Lobbying-FAQ — query-only via COPP portal

Nebraska​

Nevada​

New Hampshire​

New Jersey​

New Mexico​

New York​

North Carolina​

North Dakota​

  • Business entity search: North Dakota Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (FirstStop) — https://firststop.sos.nd.gov/search/business — query-only
  • Notary registry: North Dakota Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (FirstStop) — https://firststop.sos.nd.gov/search/notary — query-only; data list requests available
  • Charity registry: North Dakota Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://firststop.sos.nd.gov/search/charitable — query-only
  • Lobbyist registry: North Dakota Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (FirstStop) — https://firststop.sos.nd.gov/search/lobbyist — query-only; annual lists at https://firststop.sos.nd.gov/lists

Ohio​

Oklahoma​

Oregon​

Pennsylvania​

  • Business entity search: Pennsylvania Department of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (Bureau of Corporations and Charitable Organizations) — https://file.dos.pa.gov/search/business — query-only
  • Notary registry: Pennsylvania Department of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://www.notaries.pa.gov/pages/notarysearch.aspx — query-only
  • Charity registry: Pennsylvania Department of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (Bureau of Corporations and Charitable Organizations) — https://www.charities.pa.gov/ — query-only
  • Lobbyist registry: Pennsylvania Department of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. (Lobbying Disclosure) — https://www.pa.gov/services/dos/search-lobbying-disclosure-reports — query-only

Rhode Island​

South Carolina​

  • Business entity search: South Carolina Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://businessfilings.sc.gov/BusinessFiling/Entity/Search — query-only (note: officers/directors not maintained by SoS)
  • Notary registry: South Carolina Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://search.scsos.com/notaries — query-only
  • Charity registry: South Carolina Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://search.scsos.com/charities — query-only
  • Lobbyist registry: South Carolina State Ethics Commission — https://apps.sc.gov/lobbyingactivity/LAIndex.aspx — query-only

South Dakota​

Tennessee​

Texas​

Utah​

Vermont​

Virginia​

Washington​

West Virginia​

Wisconsin​

Wyoming​

  • Business entity search: Wyoming Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://wyobiz.wyo.gov/business/filingsearch.aspx — query-only
  • Notary registry: Wyoming Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. has no public online notary search. Notary info on commission only — https://sos.wyo.gov/services/notaries.aspx — Not searchable online; investigate manually
  • Charity registry: WY has NO stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-level charity registration. No public registry. Investigate manually
  • Lobbyist registry: Wyoming Secretary of StateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. — https://lobbyist.wyo.gov/Lobbyist/LobbyistSearch.aspx?mode=initial — query-only

Regulated facilities, open data portals, and parcels​

State open data portals​

License terms note: most stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. ArcGIS Hub deployments default to Esri Open Data terms (effectively public-domain or attribution); confirm per dataset. Socrata-hosted stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. portals generally publish under each stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.'s open-data terms (typically attribution-only).

Regulated establishments per state​

Each stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. lists, where available: Child care licensing, Assisted living/nursing home (stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-level), Alcohol/ABC, Cannabis (where legal & public), Auto dealer licensing.

Alabama​

Alaska​

Arizona​

Arkansas​

California​

Colorado​

Connecticut​

Delaware​

District of Columbia​

Florida​

Georgia​

Hawaii​

Idaho​

Illinois​

Indiana​

Iowa​

Kansas​

Kentucky​

Louisiana​

Maine​

Maryland​

Massachusetts​

Michigan​

Minnesota​

Mississippi​

Missouri​

Montana​

Nebraska​

Nevada​

New Hampshire​

New Jersey​

New Mexico​

New York​

North Carolina​

  • Child care: NC DHHS, Division of Child Development and Early Education — https://ncchildcaresearch.dhhs.nc.gov/ — query; bulk via NC.gov data.
  • Assisted living/nursing: NC DHHS, Division of Health Service Regulation — https://info.ncdhhs.gov/dhsr/reports.htm — facility reports + downloads.
  • Alcohol/ABC: NC ABC Commission — https://abc.nc.gov/ — license search; NC is partial control stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality..
  • Cannabis: N/A — recreational and medical cannabis illegal in North Carolina (limited hemp/CBD only).
  • Auto dealer: NC Division of Motor Vehicles, License and Theft Bureau — https://www.ncdot.gov/dmv/license-id/dealers/Pages/default.aspx — Not found — investigate manually.

North Dakota​

Ohio​

Oklahoma​

Oregon​

Pennsylvania​

Rhode Island​

South Carolina​

South Dakota​

Tennessee​

Texas​

Utah​

Vermont​

Virginia​

Washington​

West Virginia​

Wisconsin​

Wyoming​

State-aggregated parcel data​

Only statesregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. with a true statewide parcelparcelA property polygon or record carrying a situs (site) address and often a separate owner mailing address. County GIS parcel aggregations are a training source for address-point variety and situs-vs-owner divergence. aggregation (not just per-county) listed. License notes added where visible.

States NOT included here either lack stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality.-level parcelparcelA property polygon or record carrying a situs (site) address and often a separate owner mailing address. County GIS parcel aggregations are a training source for address-point variety and situs-vs-owner divergence. aggregation (county-only) or only publish partial datasets: AL, AK, AZ, CA, CO, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, MI, MS, MO, NE, NV, NJ, NM, NY, ND, OH, OK, OR, PA, SC, SD, TX, WA, WV, WI. Some of these (e.g., NJ, TX, WI, OR) maintain statewide indexes pointing to county data but do not redistribute parcelsparcelA property polygon or record carrying a situs (site) address and often a separate owner mailing address. County GIS parcel aggregations are a training source for address-point variety and situs-vs-owner divergence. as a unified statewide layerlayerOne transformer block — attention plus a feed-forward network, with normalization and residual connections — applied to every position. Stacking layers lets the model build up richer representations; Mailwoman's encoder has 6. at no cost.

Notes on bulk vs. query, and license terms​

  • ArcGIS Hub-hosted stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. portals (Esri Open Data) generally allow CSV/GeoJSON/SHP downloads with attribution; Esri terms are public-domain-equivalent by default but each dataset may override.
  • Socrata-hosted portals (data.ca.gov, data.ny.gov, opendata.maryland.gov, etc.) expose SODA APIs for query and CSV/JSON bulk export per dataset.
  • Many regulator websites (DMV dealer lookups, liquor license search portals) are query-only; bulk acquisition typically requires FOIA/public records request unless an open-data dataset is also published.
  • Cannabis: included only where stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. has a legal commercial program and publishes a licensee list. StatesregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. marked N/A either prohibit cannabis or only permit narrow medical/CBD programs without a meaningful public dispensary roster.
  • Always check per-stateregionThe first-level administrative subdivision of a country — a US state, a French region, a province. The component between country and locality. terms of use; CC-BY and CC0 are common but not universal.