Title :
Automatic Entity Relation Extraction Based on Maximum Entropy
Author :
Zhang Suxiang ; Wen Juan ; Wang Xiaojie ; Li Lei
Author_Institution :
Sch. of Inf. Eng., Beijing Univ. of Posts & Telecommun.
Abstract :
Entity relation extraction (RE) is an very important research domain in information extraction, we can regard RE as a classification problem in this paper, RE is still original study field in Chinese language now, maximum entropy (ME)-based machine learning is the first time to be used to extract entity relations between named entities from Chinese texts, Thirteen features have been designed for entity relation extraction, which includes morphology, grammar and semantic feature. The system architecture for RE has been constructed. Experiment shows that the performance is promising. So it is useful for ME-based machine learning to solve RE problem
Keywords :
grammars; learning (artificial intelligence); maximum entropy methods; natural language processing; text analysis; Chinese language; Chinese texts; automatic entity relation extraction; classification problem; feature selection; grammar; information extraction; machine learning; maximum entropy; morphology; semantic feature; Data mining; Design engineering; Entropy; Kernel; Learning systems; Machine learning; Machine learning algorithms; Natural language processing; Natural languages; Power engineering and energy; Maximum Entropy; entity relation extraction and evaluation; feature selection;
Conference_Titel :
Intelligent Systems Design and Applications, 2006. ISDA '06. Sixth International Conference on
Conference_Location :
Jinan
Print_ISBN :
0-7695-2528-8
DOI :
10.1109/ISDA.2006.115
