Title :
Exploiting web features in Chinese relation extraction
Author :
Wang Jin ; Chen Jian Ping ; Gu Xiang
Author_Institution :
Xinglin Coll., Comput. Sci. & Technol., Nantong Univ., Nantong, China
Abstract :
Relation extraction is a form of information extraction, which finds predefined relations between pairs of entities in text. A Chinese relation extraction approach exploiting web features is proposed. Four web features are extracted from the web and the Wikipedia website. Experiments on the ACE 2005 Corpus show that the web features are effective, and high-quality websites generate more effective features.
Keywords :
Web sites; information retrieval; natural language processing; ACE 2005 Corpus; Chinese relation extraction; Web features; Wikipedia Website; high-quality Websites; information extraction; Educational institutions; Electronic publishing; Encyclopedias; Feature extraction; Internet; Semantics; Wikipedia; co-occurrence; relation extraction; web feature;
Conference_Titel :
Computer Science and Automation Engineering (CSAE), 2012 IEEE International Conference on
Conference_Location :
Zhangjiajie
Print_ISBN :
978-1-4673-0088-9
DOI :
10.1109/CSAE.2012.6272550
