During the last years, the research area of taxonomy matching has made a massive progress, and many works in this area have been introduced providing significant advances to the field. The most improving attempts could highly outperform existing approaches because of two areas of innovation. The first innovation is to provide a combination of matching techniques inside a flexible matching strategy. Based on the domain, the used matching techniques and parameters can now be chosen in a flexible manner providing of course higher matching accuracy when overcoming one or multiple type(s) of heterogeneity. The second innovation is to support the matching techniques by linking to various sources of so-called background knowledge. The matching techniques, algorithms, and systems provide better quality and efficiency the more data and knowledge about the domain is provided. Using external knowledge in the form of thesaurus, linked data, or Semantic Web technologies, matching quality and efficiency could have been further increased. However, as no available survey considers recently developed matching systems according to the implemented matching strategies and utilized sources of background knowledge, a review focussing on this two criteria is required. To fill this gap, a comprehensive review is provided using this chapter.
|Title of host publication||Taxonomy Matching Using Background Knowledge|
|Subtitle of host publication||Linked Data, Semantic Web and Heterogeneous Repositories|
|Number of pages||24|
|Publication status||Published - 8 Jan 2018|