Integrity Proofs for RDF Graphs

Representing open datasets with the RDF model is becoming increasingly popular. An important aspect of this data model is that it can utilize the methods of computing cryptographic hashes to verify the integrity of RDF graphs. In this paper, we first develop a number of metrics to compare the state-of-the-art integrity proof methods and then present two new approaches to generate an integrity proof of RDF datasets: (i) semantic-based and (ii) structurebased. The semantic-based approach leverages timestamps (or other inherent notions of ordering) as an indexing key to construct a sorted Merkle tree variation, where timestamps are semantically extractable from the dataset. The structure-based approach utilizes the redundant structure of large RDF datasets to compress the dataset statements prior to generating a variation of a Merkle tree. We provide a theoretical analysis and an experimental evaluation of our two proposed methods. Compared to the Merkle and sorted Merkle tree, the semantic-based approach achieves faster querying performance for large datasets. The structure-based approach is well suited when RDF datasets contain large amounts of semantic redundancies. We also evaluate our methods’ resistance to adversarial threats.