Abstract

In the process of constructing large-scaleknowledge base, manual-based construction approach lacksefficiency as well as flexibility. Therefore, automaticallyextracting of massive knowledge from online encyclopediahas attracted attention from an increasing number ofscholars. Current research is mainly focused on theextracting of data from English online encyclopedia,whereas research about knowledge extraction from Chineseor other language data sources is rare. For such reason, thepresent paper proposes an automatic construction schemefor large-scale knowledge base based on Chinese onlineEncyclopedia. (i)In the first phase of the scheme, selfexpandedlearning is performed on the semantic relationsbetween subjects and objects among the knowledge triples.(ii)In the second phase, semantic relations between themarked attributes and their entities is predicted usingConditional Random Fields (CRFs) and Support vectormachine (SVM) classifier. A large-scale knowledge base isautomatically constructed based on the scheme, and theexperiment results indicate that the scheme possessesfeasibility and effectiveness.