Author(s): Johannes Frey, Amrapali Zaveri, Magnus Knuth, Marvin Hofer, Sebastian Hellmann
Abstract: Data quality improvement of DBpedia has been in the focus of several publications in the past years. These works cover both knowledge enrichment techniques such as type learning, taxonomy generation, interlinking, etc. and error detection strategies like property or value outlier detection, type checking, ontology constraints, unit-tests, to name just a few. The goal of DBpedia Fusion is to take advantage of Wikipedia articles in different languages, which are independently maintained by authors from the individual language chapters. In this paper we define a set of quality metrics and evaluate them for Wikidata and DBpedia datasets of several language chapters. Moreover, we show that a quality-driven knowledge fusion approach of these datasets increases data richness as well as correctness.
Keywords: Data Fusion; Quality Assessment; DBpedia; Linked Data