INTENT

The INterlinear Text ENrichment Toolkit. This readme will describe the aims of the INTENT package, as well as a quickstart guide to get up and running.

Project Goals

The INTENT package is intended to take instances of Interlinear Glossed Text (IGT) found in places such as the Online Databse of Interlinear Text (ODIN)¹ and use a variety of methods to enrich the instances. Currently, INTENT:

• Adds alignment information to IGT instances
• Adds POS Tag information to IGT instances

And INTENT will soon:

• Add dependency structures to IGT

Dependencies

Currently, INTENT is written in Python 3, and depends on several external python modules and other tools.

###Python Modules ###

• XIGT
• NLTK
• The lxml XML library

These modules must be either installed or on the PYTHONPATH variable when running INTENT. pip is strongly reccomended for installing NLTK and lxml.

###Other Tools ###

• MALLET (MAchine Learning for LanguagE Toolkit)
• Stanford Log-linear Part-Of-Speech Tagger
• mgiza++

Footnotes

1. Xia, F., Lewis, W. D., Goodman, M. W., & Crowgey, J. (2014). Enriching ODIN. Presented at the Proceedings of the Ninth International Conference on Language Resources and Evaluation, Reykjavik, Iceland. ↩