The method of Constructing Chinese Knowledge Base based on open source English Knowledge Base

Abstract

Building a large-scale knowledge base is a hot spot in the current research of big data, and it is the basis of Internet search, Question Answering (Q&A), Machine Translation (MT) etc. At present, it has formed serval mature English knowledge base, such as YAGO, Freebase, DBpedia, etc. but we still lack a large-scale Chinese knowledge base. This paper presents an approach to construct the Chinese knowledge base by using the automatic translation technique and the method of building the Chinese knowledge base with the source of English knowledge base. And proposing a quality calibration method based on the search engine. In order to verify the above method, we take the YAGO English knowledge base as an example, the formation of Chinese Knowledge most quality of knowledge is up to 95%, and a few low quality data can be discarded after being calibrated. The method is an important supplement to building a large-scale knowledge base by using text information extraction, crowdsourcing and other methods.