IEPY is a framework for doing information extraction on unstructured documents. It uses partially supervised machine learning techniques (i.e., there's a human helping the application, but the application generalizes what the human does and learns).

Typical applications have a set of text documents as input (for example a Wiki, or a database of patent applications). Those documents refer to some entities of different kinds (for example “Albert Einstein” may be an entity of kind “person” and “physicist” is an entity of kind “profession”). The application defines the relevant entity kinds and the relations between those kinds to extract (for example “person HAD-PROFESSION profession”). The other input required is a set of seed facts, which are facts known to be true, for example “Albert Einstein HAD-PROFESSION physicist”.

From that information, a IEPY application is able to find other examples of the relation in the input documents, between the provided entities (example: “Albert Einstein HAD-PROFESSION patent clerk”) or even between other unrelated entities (example: “Ernest Hemingway HAD-PROFESSION writer”). These extracted facts are also tagged with fragments of the original documents that are evidence of the fact (example: “In late 1919 Ernest Hemingway began as a freelancer, staff writer, and foreign correspondent for the Toronto Star Weekly.”). During the extraction process, a person helps the system by replying yes/no to questions of the form “Does this text fragment reflect this other fact?”

Please check docs/install.rst

Available at http://iepy.readthedocs.org/en/latest/.

IEPY is © 2014 Machinalis in collaboration with the NLP Group at UNC-FaMAF. Its primary authors are:

• Rafael Carrascosa <rcarrascosa@machinalis.com> (rafacarrascosa at github)
• Franco M. Luque <francolq@famaf.unc.edu.ar> (francolq at github)
• Javier Mansilla <jmansilla@machinalis.com> (jmansilla at github)
• Daniel Moisset <dmoisset@machinalis.com> (dmoisset at github)

You can follow the development of this project and report issues at http://github.com/machinalis/iepy

This project has a BSD license, as stated in the LICENSE file.

No stable releases yet. Coming soon.

The project is currently working, it has good testing coverage and a working example. We're still missing some API cleanup, documentation, packaging and a couple of large bugfixes.