Skip to main content
Help us improve Python packaging – donate today!

Hidden alignment conditional random field, a discriminative string edit distance

Project Description

https://travis-ci.org/datamade/pyhacrf.svg?branch=master https://ci.appveyor.com/api/projects/status/kibqrd7wnsk2ilpf/branch/master?svg=true

Hidden alignment conditional random field for classifying string pairs - a learnable edit distance.

This package aims to implement the HACRF machine learning model with a sklearn-like interface. It includes ways to fit a model to training examples and score new example.

The model takes string pairs as input and classify them into any number of classes. In McCallum’s original paper the model was applied to the database deduplication problem. Each database entry was paired with every other entry and the model then classified whether the pair was a ‘match’ or a ‘mismatch’ based on training examples of matches and mismatches.

I also tried to use it as learnable string edit distance for normalizing noisy text. See A Conditional Random Field for Discriminatively-trained Finite-state String Edit Distance by McCallum, Bellare, and Pereira, and the report Conditional Random Fields for Noisy text normalisation by Dirko Coetsee.

Example

from pyhacrf import StringPairFeatureExtractor, Hacrf

training_X = [('helloooo', 'hello'), # Matching examples
              ('h0me', 'home'),
              ('krazii', 'crazy'),
              ('non matching string example', 'no really'), # Non-matching examples
              ('and another one', 'yep')]
training_y = ['match',
              'match',
              'match',
              'non-match',
              'non-match']

# Extract features
feature_extractor = StringPairFeatureExtractor(match=True, numeric=True)
training_X_extracted = feature_extractor.fit_transform(training_X)

# Train model
model = Hacrf(l2_regularization=1.0)
model.fit(training_X_extracted, training_y)

# Evaluate
from sklearn.metrics import confusion_matrix
predictions = model.predict(training_X_extracted)

print(confusion_matrix(training_y, predictions))
> [[0 3]
>  [2 0]]

print(model.predict_proba(training_X_extracted))
> [[ 0.94914812  0.05085188]
>  [ 0.92397711  0.07602289]
>  [ 0.86756034  0.13243966]
>  [ 0.05438812  0.94561188]
>  [ 0.02641275  0.97358725]]

Dependencies

This package depends on numpy. The LBFGS optimizer in pylbfgs is used, but alternative optimizers can be passed.

Install

Install by running:

python setup.py install

or from pypi:

pip install pyhacrf

Developing

Clone from repository, then

pip install -r requirements.txt
cython pyhacrf/*.pyx
python setup.py install

To deploy to pypi, make sure you have compiled the *.pyx files to *.c

Release history Release notifications

This version
History Node

0.2.1

History Node

0.2.0

History Node

0.1.2

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
pyhacrf_datamade-0.2.1-cp27-cp27m-macosx_10_11_x86_64.whl (208.1 kB) Copy SHA256 hash SHA256 Wheel cp27 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp27-cp27m-manylinux1_i686.whl (664.8 kB) Copy SHA256 hash SHA256 Wheel cp27 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp27-cp27m-manylinux1_x86_64.whl (712.9 kB) Copy SHA256 hash SHA256 Wheel cp27 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp27-cp27mu-manylinux1_i686.whl (664.7 kB) Copy SHA256 hash SHA256 Wheel cp27 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp27-cp27mu-manylinux1_x86_64.whl (713.0 kB) Copy SHA256 hash SHA256 Wheel cp27 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp27-cp27m-win32.whl (148.7 kB) Copy SHA256 hash SHA256 Wheel cp27 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp27-cp27m-win_amd64.whl (177.8 kB) Copy SHA256 hash SHA256 Wheel cp27 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp34-cp34m-manylinux1_i686.whl (685.6 kB) Copy SHA256 hash SHA256 Wheel cp34 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp34-cp34m-manylinux1_x86_64.whl (729.7 kB) Copy SHA256 hash SHA256 Wheel cp34 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp34-cp34m-win32.whl (143.4 kB) Copy SHA256 hash SHA256 Wheel cp34 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp34-cp34m-win_amd64.whl (164.2 kB) Copy SHA256 hash SHA256 Wheel cp34 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp35-cp35m-manylinux1_i686.whl (676.0 kB) Copy SHA256 hash SHA256 Wheel cp35 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp35-cp35m-manylinux1_x86_64.whl (722.3 kB) Copy SHA256 hash SHA256 Wheel cp35 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp36-cp36m-macosx_10_11_x86_64.whl (200.8 kB) Copy SHA256 hash SHA256 Wheel cp36 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp36-cp36m-manylinux1_i686.whl (682.4 kB) Copy SHA256 hash SHA256 Wheel cp36 Mar 27, 2017
pyhacrf_datamade-0.2.1-cp36-cp36m-manylinux1_x86_64.whl (727.8 kB) Copy SHA256 hash SHA256 Wheel cp36 Mar 27, 2017
pyhacrf_datamade-0.2.1-py2.7-win32.egg (146.9 kB) Copy SHA256 hash SHA256 Egg 2.7 Mar 27, 2017
pyhacrf_datamade-0.2.1-py2.7-win-amd64.egg (176.0 kB) Copy SHA256 hash SHA256 Egg 2.7 Mar 27, 2017
pyhacrf_datamade-0.2.1-py3.4-win32.egg (141.6 kB) Copy SHA256 hash SHA256 Egg 3.4 Mar 27, 2017
pyhacrf_datamade-0.2.1-py3.4-win-amd64.egg (162.3 kB) Copy SHA256 hash SHA256 Egg 3.4 Mar 27, 2017
pyhacrf-datamade-0.2.1.tar.gz (255.8 kB) Copy SHA256 hash SHA256 Source None Mar 27, 2017

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging CloudAMQP CloudAMQP RabbitMQ AWS AWS Cloud computing Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page