Xigt is both a data model for representing annotations of natural language data, as well as the canonical XML serialization of this data model. The kind of data Xigt is intended to encode is primarily interlinear glossed text, although other kinds of data are possible.

The main features of Xigt are:

• simple, predictable data structure
• alignment across tiers by reference via alignment expressions
• lean but versatile schema for standard data
• easily extensible for other kinds of data

• Introduction - the Xigt data model and formats
• Schemata - the default schema and schema extensions
• API Reference - the Python API for working with Xigt corpora
• Tutorials - practical working examples and instructions

• The README
• Some examples
• INTENT - IGT enrichment using Xigt
• A Xigt Editor for Windows