D-RDF: Dynamic Resource Description Framework.

摘要

Semantic Web is described as the Web of Data, as opposed to the World Wide Web which is a Web of Documents. As research in the field of Semantic Web is gaining momentum, the focus is shifting on the effective representation of the data that constitutes the Semantic Web. RDF or the Resource Description Framework is the W3C standardized language for describing the semantics of the data and hence, sharing its meaning across applications. In RDF, all entities are modeled as resources and facts about these resources are asserted in terms of properties and their values.;In this thesis, we propose a computation model for RDF called Dynamic RDF or D-RDF. While RDF models are restricted to describing resources in terms of their assertive or static properties, D-RDF is a generalization of RDF where there exist the assertive properties of the resources along with certain dynamic properties that operate on the values of existing properties and infer new data. In such a model, the information carries the semantics with it in the form of computing methods. In other words, whereas RDF represents semantically enhanced data, D-RDF represents both data and the programs that operate on the data. This design ensures that when the D-RDF model is processed, the dynamic properties would operate on the current values of the base data and hence, the values of the dynamic properties will always be consistent with changes that occur on that data. Hence, we develop a model of context-sensitive semantics and implement an interpretation engine for this language. The power of D-RDF is demonstrated by implementing use-cases of varying levels of complexity that highlight how highly customized data models can be constructed with D-RDF to represent information in a form that does not already exist.