Provenance: The Missing Component of the Semantic Web for Privacy and Trust

Data on the Semantic Web currently does not have any standardized or any de-facto agreed upon way to exhibit provenance information, yet provenance is the foundation for any reasonable model of privacy and trust. Yet, currently every RDF triple does not have any coherent way of storing provenance information on the Semantic Web. We present the hypothesis that provenance is by far the most important data needed on the Semantic Web for privacy and trust, and review previous work in database systems on provenance. We put forward the concept that the three main provenances operators (insertion, deletion, and copy) from provenance work in database systems can be used on the Semantic Web. Furthermore, we hypothesize that such information naturally should be stored in or using the name URI of named graphs. We show that such an approach can help solve practical issues of privacy and trust in social networks using a real-world example.