Foundations of data interoperability on the web

In this paper, when we use the term ontology, we are primarily referring to linked data in the form of RDF(S). The problem of ontology mapping has attracted considerable attention over the last few years, as the deployment of ontologies is increasing with the advent of the Web of Data. We identify two sharply distinct goals for ontology mapping, based on real-world use cases. These goals are: (i) ontology development, and (ii) facilitating interoperability. We systematically analyze the goals, side-by-side, and contrast them for the first time. Our analysis demonstrates the implications of the goals on ontology mapping and mapping representation. Many studies on ontology mapping have focused on ontology merging. Ontology merging is an ontology development task (goal i). With the increase in the number of web-based information systems that utilize ontologies, the need for facilitating interoperability between these systems is becoming more visible (goal ii).