MMAX2 is a highly customizable annotation tool
for creating, browsing, visualizing and querying linguistic annotations
on multiple levels. It mainly supports the manual annotation process,
but it also easily integrates annotations from other sources, like e.g.
automatic pre-processing components. MMAX2 was designed to allow the creation and maintenance of annotated corpora which are highly usable and re-usable. Why?
Annotation is one of the key tasks in empirically oriented linguistics.
Manually annotated data is important for both research and development,
as basis for computer-aided exploration of linguistic phenomena and
their interaction, and for building and evaluating NLP components.
Since manual annotation is time-consuming and expensive, the usability
of annotations should not be limited to the task they were originally
created for. How?
MMAX2 uses a flexible 'stand-off' annotation model transparently
implemented in XML. It uses a single file format for all levels and
thus allows for the trouble-free coexistence of arbitrarily many
annotation levels in a single document. This also supports distributed
annotation, with each annotation task being worked at by a specialized
group. The tool's GUI offers advanced methods for rendering and
interacting with information from different annotation levels.
For more information, including download, see the official MMAX2 web site at http://mmax2.net.
page last modified: 22.12.2011,15:35
Project Manager
Prof. Dr. Michael Strube Email:
Phone: +49 (0)6221 - 533 - 243 Fax: +49 (0)6221 - 533 - 298 more >>