文献类型：期刊文献

英文题名：Evaluating and comparing web-scale extracted knowledge bases in Chinese and english

作者：Ruan, Tong[1]; Dong, Xu[1]; Wang, Haofen[1]; Li, Yang[1]

机构：[1] Department of Computer Science and Engineering, East China University of Science and Technology, Shanghai, 200237, China

年份：2016

卷号：9544

起止页码：167

外文期刊名：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

收录：EI(收录号：20161302169491)

语种：英文

摘要：DBpedia and YAGO are the two main data sources serving as the hub of Linking Open Data (LOD), and they both contain Chinese data. Zhishi.me and SSCO extract Chinese knowledge from Wikipedia and other Chinese EncyclopedicWeb sites like Baidu-Baike and Hudong- Baike. The quality of these Knowledge Bases (KBs) are not well investigated while their qualities are key to smart applications. In this paper, we evaluate three large Chinese KBs including DBpedia Chinese, zhishi.me and SSCO, and further compare them with English KBs. Since traditional methods on evaluating Web ontology can not be easily adapted to web-scale extracted KBs, we design two metric sets considering Richness and Correctness based on a quasi-formal conceptual representation to measure and compare these KBs. We also design a novel metric set on overlapped instances of different KBs to make the metric results comparable. Finally, we employ random sampling to reduce human efforts for assessing the correctness. The findings in these KBs give a detailed status report of the current situation of extracted KBs in both Chinese and English. ? Springer International Publishing Switzerland 2016.