Publikation

Adapting SproUT to processing Baltic and Slavonic Languages

Witold Drozdzynski, Petr Homola, Jakub Piskorski, Vytautas Zinkevicius

In: Proceedings of the Workshop Information Extraction for Slavonic Languages, held in conjunction with the Conference Recent Advances in Natural Language Processing. Workshop Information Extraction for Slavonic Languages September 10-12 Borovets Bouvet Island 2003.

Abstrakt

This paper focuses on presenting an initial effort for porting SProUT - a novel general purpose IE platform, to processing Baltic and Slavonic languages. We describe the system, characterize the mentioned language groups and discuss the process of developing named-entity and chunk grammars, which are crucial for solving information extraction tasks.

homola_baltslavir.pdf (pdf, 240 KB )

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence