Affiliation:
1. Beijing Information Science and Technology University
Abstract
In this paper, we have designed and realized a efficient full-text retrieval system for the basic annotation People's Daily Corpus based on the inverted index technology. According to the characteristics of the basic annotation People’s Daily Corpus data, we have analyzed the methods and strategies of system implementing thoroughly. On the basis of comparing the various schemes, we have put forward to the three levels index structure of Chinese character, word and address set, and given the design approach of each level index dictionary structure. After converting the unstructured People’s Daily corpus into index structured data, we realized the full-text search algorithm correspond to the proposed index structure. Experimental results show that the proposed search algorithm has achieved the target of "ten millions Chinese characters, response in a second", improved the speed of the People's Daily Corpus full-text search.
Publisher
Trans Tech Publications, Ltd.
Reference4 articles.
1. Long Haiyan, Huo Yuejun. Database query optimization [J]. Computer Knowledge and Technology, 2003(1): 8-9. (Chinese).
2. Lang Xiaowei, Wang Shenkang. Research and development of full-text retrieval system based on Lucene[J]. Computer Engineering, 2006(4): 94-99. (Chinese).
3. Su YingTan, Guo Xianyong, Jin xin. Chinese text retrieval system based Lucene [J]. Computer Engineering, 2007(23): 94-96. (Chinese).
4. Chen Aitao , He Jianzhang , Xu Liangjie , Fredric C. Gey , Jason Meggs, Chinese text retrieval without using a dictionary, Proceedings of the 20th annual international ACM SIGIR conference on Research and development in information retrieval, 1997, July 27-31: 42-49.
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献