Machine Knowledge: Creation and Curation of Comprehensive Knowledge\n Bases

Abstract

Equipping machines with comprehensive knowledge of the world's entities and\ntheir relationships has been a long-standing goal of AI. Over the last decade,\nlarge-scale knowledge bases, also known as knowledge graphs, have been\nautomatically constructed from web contents and text sources, and have become a\nkey asset for search engines. This machine knowledge can be harnessed to\nsemantically interpret textual phrases in news, social media and web tables,\nand contributes to question answering, natural language processing and data\nanalytics. This article surveys fundamental concepts and practical methods for\ncreating and curating large knowledge bases. It covers models and methods for\ndiscovering and canonicalizing entities and their semantic types and organizing\nthem into clean taxonomies. On top of this, the article discusses the automatic\nextraction of entity-centric properties. To support the long-term life-cycle\nand the quality assurance of machine knowledge, the article presents methods\nfor constructing open schemas and for knowledge curation. Case studies on\nacademic projects and industrial knowledge graphs complement the survey of\nconcepts and methods.\n