2026-10-12T21:56:59Z https://www.scidb.cn/oai

10.11922/sciencedb.j00104.00041 2020-10-16T16:42:27Z

2020-10-16 CN-DBpedia2: An Extraction and Verification Framework for Enriching Chinese Encyclopedia Knowledge Base doi:10.11922/sciencedb.j00104.00041 en Three tables and eight figures of this paper. Table 1 shows the F1 scores on slot filling benchmark TAC data set (development (dev) set: data from 2012/2013, evaluation (eval) set: data from 2014). Table 2 shows fact types in CN-DBpedia2, as well as changes compared to CN-DBpedia. Table 3 presents the top 20 concepts and the number of entities included in CN-DBpedia2, as well as changes compared to CN-DBpedia. Figure 1 shows system architecture of CN-DBpedia2. Note: It is an extension of CN-DBpedia, and the dotted components are new features. Figure 2 presents the detail of enrichment component in CN-DBpedia2. Figure 3 is the framework of METIC, a multi-instance method for entity typing from unstructured features. Figure 4 shows a neural network model for mention typing. Figure 5 is a framework of METIH, a multi-instance method for entity typing from heterogeneous features. Figure 6 shows network structures of the multiple-predicate extractor. Figure 7 presents an example of CAPTCHA. Note: In this case, the answer is “overlooks inquiry”. Figure 8 is an example of our rcCAPTCHA system. Note: The word marked in yellow is the correct answer. Knowledge graph; Entity typing; Slot filling; Information extraction; Crowdsourcing Bo Xu Jiaqing Liang Chenhao Xie Bin Liang Lihan Chen Yanghua Xiao PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset http://www.doi.org/10.1162/dint_a_00017 Science Data Bank