Adapting A Robust Multi-Genre Named Entity System for Automatic Content Extraction

Maynard, Dr Diana and Tablan, Mr Valentin and Bontcheva, Dr. Kalina and Dimitrov, Mr Marin (2002) Adapting A Robust Multi-Genre Named Entity System for Automatic Content Extraction. Proceedings of the Tenth International Conference on Artificial Intelligence: Methodology, Systems, Applications (AIMSA 2002),, Sozopol, Bulgaria.

Full text available as:

PDF - Requires Adobe Acrobat Reader or other PDF viewer.

Many current information extraction systems tend to be designed with particular applications and domains in mind. With the increasing need for robust language engineering tools which can handle a variety of language processing demands, we have used the GATE architecture to design MUSE - a system for named entity recognition and related tasks. In this paper, we address the issue of how this general-purpose system can be adapted for particular applications with minimal time and effort, and how the set of resources used can be adapted dynamically and automatically. We focus specifically on the challenges of the ACE (Automatic Content Extraction) entity detection and tracking task, and preliminary results show promising figures.

Contact the site administrator at: hg@ecs.soton.ac.uk