Entity matching (EM) is a critical problem and will become even more critical. Lot of EM research, but has focused mostly on matching algorithms. Focus mostly on maximizing some performance factors such as accuracy, time, money, etc. Currently there is very little help for users. For many of these steps, there is little or no work, so no solution. Even when there are solutions, there may be no tools. Even when there are tools, user is faced with a wide variety of tools, one for each step, so user is left trying to move among the tools, stitching them together, deciphering their different data formats, import/export commands. too cumbersome and difficult. So often users just give up and roll their own ad-hoc solutions instead. This is obviously a major bottleneck that prevents the deployment of matching in practice. To solve the above problems, py_entitymatching aims to provide a set of commands that help the user to
- Come up with a EM workflow
- Iterate and debugmatcher the workflow
- Deploy it in production
The package is free, open-source, and BSD-licensed.
- Project Homepage: https://sites.google.com/site/anhaidgroup/projects/py_entitymatching
- Code repository: https://github.com/anhaidgroup/py_entitymatching
- Issue Tracker: https://github.com/anhaidgroup/py_entitymatching/issues
- Mailing List: https://groups.google.com/forum/#!forum/py_entitymatching
py_entitymatching has been tested on Linux, OS X and Windows.