End-to-end deduplication solution
Project description
DedupliPy
Deduplication is the task to combine different representations of the same real world entity. This package implements deduplication using active learning. Active learning allows for rapid training without having to provide a large, manually labelled dataset.
DedupliPy is an end-to-end solution with advantages over existing solutions:
- active learning; no large manually labelled dataset required
- during active learning, the user gets notified when the model converged and training may be finished
- works out of the box, advanced users can choose settings as desired (custom blocking rules, custom metrics, interaction features)
Developed by Frits Hermans
Documentation
Documentation can be found here
Installation
Normal installation
Install directly from Pypi:
pip install deduplipy
Install to contribute
Clone this Github repo and install in editable mode:
python -m pip install -e ".[dev]"
python setup.py develop
Usage
Apply deduplication your Pandas dataframe df
as follows:
myDedupliPy = Deduplicator(col_names=['name', 'address'])
myDedupliPy.fit(df)
This will start the interactive learning session in which you provide input on whether a pair is a match (y) or not (n). During active learning you will get the message that training may be finished once algorithm training has converged. Predictions on (new) data are obtained as follows:
result = myDedupliPy.predict(df)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for DedupliPy-0.6.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 045dce199b2662d4dfb0f44bbc753dac430259de8db96e4837846498369a40a7 |
|
MD5 | 3d2d800c2b78cd81e1bfe269daa10912 |
|
BLAKE2b-256 | 681bb53038609e42c5e707a52bfff3085d07a8c2bd1b55803a7cf25438c8c658 |