Mining Wikipedia for Discovering Multilingual Definitions on the Web

Alejandro Figueroa

In: Proceedings of the 4th International Conference on Semantics, Knowledge and Grid. International Conference on Semantics, Knowledge and Grid (SKG-2008), 4th, December 3-5, Beijing, Pekin, China, Pages 125-132, IEEE, 12/2008.


Ml-DfWebQA is a multilingual definition question answering system (QAS) that extracts answers to definition queries from the short descriptions of web-sites returned by search engines, called web snippets. These answers are discriminated on the ground of lexico-syntactic regularities mined from multilingual resources supplied by Wikipedia. Results support that these regularities serve to significantly strengthen the answering process. In addition, Ml-DfWebQA increases the robustness of multilingual definition QASs by making use of aliases found in Wikipedia.

skg2008_v3.pdf (pdf, 290 KB )

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence