deepparse BPEmb
A deepparse address parser (Seq2Seq, without attention mechanism) using the multilingual BPEmb subword embeddings (100,000 merges, 300 dimensions), embedded per word by a bidirectional LSTM.
from deepparse.parser import AddressParser
address_parser = AddressParser(model_type="bpemb")
address_parser("350 rue des Lilas Ouest Québec Québec G1L 1B6")
Architecture
One-layer LSTM encoder and decoder, hidden size 1,024, BPEmb embeddings of 300 dimensions. The
model predicts 8 tags (StreetNumber, StreetName, Unit, Municipality, Province, PostalCode, Orientation
and GeneralDelivery) and the end-of-sequence tag.
Training
Trained in 2020 on the dataset described in the deepparse-address-data repository (addresses of 20 countries, built from the libpostal data). See Leveraging Subword Embeddings for Multinational Address Parsing and Multinational Address Parsing: A Zero-Shot Evaluation.
Results
- 20 training countries, clean addresses: 98.71% mean per-country accuracy (96.97% to 99.99%).
- 20 training countries, incomplete addresses: 98.24% (95.20% to 99.89%).
- 41 zero-shot countries: 70.91% (34.97% to 94.46%).
Per-country tables: the Results page of the documentation.
Limitations
- On the test split of the newer worldwide-addresses data (same 20 countries, clean and incomplete addresses, 260,000 addresses), this model reaches 94.66% per-word accuracy (measured in November 2026). It is weaker on incomplete addresses (e.g. a postal code or a municipality alone) than on complete ones.
- In practice, it rarely predicts the
OrientationandGeneralDeliverytags: in November 2026, it predicted neither on 12,000 addresses of the worldwide-addresses test split nor on typical cases such asOuest,NWorPO Box.
Citation
@misc{yassine2020leveraging,
title={{Leveraging Subword Embeddings for Multinational Address Parsing}},
author={Marouane Yassine and David Beauchemin and François Laviolette and Luc Lamontagne},
year={2020},
eprint={2006.16152},
archivePrefix={arXiv}
}