Linked Data Services and Semantics-enabled Mashup (Chapter 11)

The Web of Linked Data can be seen as a global database, where resources are identified through URIs, are self-described (by means of the URI dereferencing mechanism), and are globally connected through RDF links. Recently, research on the Web of Linked Data has been devoted to the study of models and languages to add functionalities to the Web of Linked Data by means of Linked Data services. In this chapter, we propose a new approach for “serving up” Linked Data by semantics-enabled mashup of existing Web APIs. A basic semantics-enabled model of Web APIs is proposed. The model has been designed (1) to support providers who publish new Web APIs used to access the Web of Linked Data, (2) to support the Web designer who aims at exploring and selecting available Web APIs to build or maintain a Web mashup, and (3) to make it available on top of the Web of Linked Data. Based on the proposed model, we define automated matching techniques apt to establish semantic linksamong Linked Web APIs. The model and the techniques we propose adhere to the Linked Data principles for publishing new resources on the Web of data: (1) HTTP URIs are used to identify published Web APIs, (2) useful RDF information are provided on the Web API description when someone looks up a Web API through its URI, and (3) semantic links are set to relate Web APIs to each other, thus enabling easy development and sharing of new Web mashups.