Title :
Question classification oriented Chinese Framenet feature selection
Author :
Geng Yan ; Wang Wenjing
Author_Institution :
Dept. of Inf. Eng., Shanxi Univ., Taiyuan, China
Abstract :
For feature selection of the classification in Restricted-domain Question Answering System, Chinese Net framework full advantage of this Chinese characteristics in terms of semantic expression, this paper presents feature selection called strong class information words (SCIW). First, selecting five kinds of Chinese Framenet features, and using SCIW to select features. According to each category´s classification precision of features, sorts them. Through the experiment of the combination of the features, we select the combination of features. It improved the accuracy of classification, reaching the feature reduction.
Keywords :
classification; natural language processing; question answering (information retrieval); semantic Web; Chinese Net framework; Chinese characteristics; Chinese framenet features; SCIW; category classification precision; classification accuracy; feature reduction; question classification oriented Chinese framenet feature selection; restricted-domain question answering system; semantic expression; strong class information words; Accuracy; Computers; Educational institutions; Feature extraction; Iron; Semantics; Syntactics; Chinese framenet; feature selection; question classification;
Conference_Titel :
Computer Science & Education (ICCSE), 2011 6th International Conference on
Conference_Location :
Singapore
Print_ISBN :
978-1-4244-9717-1
DOI :
10.1109/ICCSE.2011.6028622