Temporal Extensions to RDF

Abstract

The Semantic Web aims at building a foundation of semantic-based data models and languages for not only manipulating data and knowledge, but also in decision making by machines. Naturally, time-varying data and knowledge are required in Semantic Web applications to incorporate time and further reason about it. However, the original specifications of RDF and OWL do not include constructs for handling time-varying data and knowledge. For simplicity, RDF model is confined to binary predicates, hence some form of reification is needed to represent higher-arity predicates. To this date, there are many proposals extending RDF and OWL for handling temporal data and knowledge. They all focus on the valid time. In this paper, we examine each of these proposals and develop a taxonomy to classify them according to the form of reification employed: explicit reification or implicit reification. The implicit reification proposals are further divided into three sub-categories according to semantic constructs they use. Some of these proposals stay compliant to the RDF and OWL standards whereas others add new constructs to RDF model and SPARQL query language. Additionally, we compare these proposed models with respect to characteristics, such as their syntax and semantics, their compliance to RDF and OWL specifications, their need for additional objects, etc. The comparison provides a useful guideline for the researchers and practitioners of the Semantic Web in managing temporal data and knowledge.