Data linking over RDF knowledge graphs: A survey

摘要: Instance matching (IM) is the process of instances across Knowledge Bases (KBs) that refer to same real‐world object (eg, person in two different KBs). Several approaches literature were developed perform this using algorithmic techniques and search strategies. In article, we aim provide rationale for IM survey existing algorithms performing task. We begin by identifying importance such a define it formally. also new classification these depending on “source evidence,” which can be considered as context information integrated explicitly or implicitly process. discuss state‐of‐the‐art methods regarding information. We, furthermore, describe compare relation several criteria. Such comprehensive comparative study constitutes an asset guide future research IM.