Using the Semantic Annotation of Web Table Data for Knowledge Base Construction

Abstract

Knowledge bases are one of the main elements of intelligent systems. In that, knowledge base engineering is traditionally considered as a "bottleneck" in the design of such systems, and it is a deterrent factor of their widespread use. Computer-aided construction of knowledge base that employs various information sources is a promising area of scientific research. Web tables may be one of such sources, and they are among the most accessible and common ways for representing and storing tabular information. In this paper, we propose to automate ontological knowledge base engineering by using the semantic annotation of data from web tables and present an approach to filling the ontological knowledge base with specific entities (facts) extracted from web tables. This approach is implemented in the form of a tool. The tool has been used to solve the problem of forming domain knowledge graphs for the TALISMAN framework.