Paper: A Machine Learning Approach To German Pronoun Resolution

ACL ID P04-2010
Title A Machine Learning Approach To German Pronoun Resolution
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2004
Authors

This paper presents a novel ensemble learning approach to resolving German pronouns. Boosting, the method in question, combines the moderately ac- curate hypotheses of several classifiers to form a highly accurate one. Exper- iments show that this approach is su- perior to a single decision-tree classi- fier. Furthermore, we present a stan- dalone system that resolves pronouns in unannotated text by using a fully auto- matic sequence of preprocessing mod- ules that mimics the manual annotation process. Although the system performs well within a limited textual domain, further research is needed to make it effective for open-domain question an- swering and text summarisation.