Documentation for the NITE XML Toolkit

Jean Carletta

Stefan Evert

Jonathan Kilgour

Craig Nicol

Dennis Reidsma

Judy Robertson

Holger Voormann

Revision History
Revision 0.311 June 09
Revision 0.228 March 07
Revision 0.112 Jan 07

Table of Contents

Drafts before v1.0
A Basic Introduction to the NITE XML Toolkit
Downloading and Using NXT
Prerequisites
Getting Started
Setting the CLASSPATH
How to Play Media signals in NXT
Programmatic Controls for NXT
Compiling from Source and Running the Test Suites
Data
The NITE Object Model
The NITE Data Set Model
Data Storage
Metadata
Dependency Structures
Data validation
Data Set Design
Data Builds
The NXT Query Language (NQL)
General structure of a simple query
Property tests
Comments
Structural relations
Temporal relations
Quantifier
Query results
Complex queries
Known Problems
Helpful hints
Related documentation
Analysis
Command line tools for data analysis
Projecting Images Of Annotations
Reliability Testing
Graphical user interfaces
Preliminaries
Generic tools that work on any data
Configurable end user coding tools
Libraries to support GUI authoring
Using NXT in conjunction with other tools
Recording Signals
Transcription
Importing Data into NXT
Exporting Data from NXT into Other Tools
Knitting and Unknitting NXT Data Files
General Approaches to Processing NXT Data
Manipulating media files
A. FAQ
B. How To Use Metadata
What metadata files do
What metadata files look like
Metadata examples
Using Metadata to validate data
Validation limitations
C. Comparison to other efforts
Annotation Graph Toolkit (AGTK)
ATLAS
MMAX
EMU
Others
Relationship to the Text Encoding Initiative
Summary of Answer
Data without crossing hierarchies or timing
Crossing hierarchies
Timing data
Standardized GUIs
Other frameworks
D. Information and Further Reading
NXT's history and funding
Technical Documents
Documentation for Programmers
Academic publications
Bibliography

Drafts before v1.0

In October 2006, we decided to move over NXT documentation from being completely web-based to being written in DocBook so that we can generate HTML, JavaDoc, and PDF at will. We are rewriting much much of the documentation at the same time. Versions of the documentation numbered before v1.0 are incomplete, although the outline gives some idea of our intentions for it. In this version, version 0.1, not all of the information has been checked for accuracy yet. The most likely difficulties concern the following areas: corpus resources, ontologies, and object sets; validation; incomplete description of data set concepts. In addition, not all the formatting works, and the query reference manual has not been fully converted over to DocBook, so it is incomplete and hard to read in this version.