Matching UK addresses using Splink

High performance address matching using a pre-trained Splink model.

Assuming you have two duckdb dataframes in this format:

Match them with:

from uk_address_matcher.cleaning_pipelines import (
    clean_data_using_precomputed_rel_tok_freq,
)
from uk_address_matcher.splink_model import _performance_predict

df_1_c = clean_data_using_precomputed_rel_tok_freq(df_1, con=con)
df_2_c = clean_data_using_precomputed_rel_tok_freq(df_2, con=con)

linker, predictions = _performance_predict(
    df_addresses_to_match=df_1_c,
    df_addresses_to_search_within=df_2_c,
    con=con,
    match_weight_threshold=-10,
    output_all_cols=True,
    include_full_postcode_block=True,
)

Initial tests suggest you can match ~ 1,000 addresses per second against a list of 30 million addresses on a laptop.

Refer to the example, which has detailed comments, for how to match your data.

See an example of comparing two addresses to get a sense of what it does/how it scores

Run an interactive example in your browser:

Match 5,000 FHRS records to 21,952 companies house records in < 10 seconds.

Investigate and understand how the model works