Machine learning techniques are widely used today for a variety of tasks. In this contest, ML techniques will be applied to Part-of-Speech (POS) tagging and chunking of Indian languages. These are usually essential tasks for any NLP application. This shared task aims to collectively explore various machine learning techniques (example, Hidden Markov Models) to solve the problem of POS tagging and chunking for Indian languages.

Competition:
In the competition, training data will be provided to the contestants. It will consist of annotated text. The contestants will have to train their systems on the data. Test data will be made available to them to improve and refine their system. The final contest will be held in January, 2007 with new test data. A workshop will be held to allow the shortlisted contestants to present their techniques and results.
The contestants may also tag data with the given tag set and submit to the contest. Those who do not submit may still participate. They will, however, be provided with reduced size of training data.

Resources:
*Tools which implement various machine learning techniques are available.
*Data submitted by all participants would be available to all the participants who submit data.
For any clarifications, please contact NLP Tools Contest Chair on icon07contest@iiit.ac.in