2017/08/30 by Tobias Kuhn, Kuhn, Tobias, Egon Willighagen +11
Biochemistry, Genetics and Molecular Biology · Computer Science · #Biomedical Text Mining and Ontologies #Digital Libraries (cs.DL) #FOS: Computer and information sciences #Library Science and Information Systems #Semantic Web and Ontologies #cs.DL
paper · pdf · doi:10.48550/arxiv.1708.09193
In Proceedings of the 16th International Semantic Web Conference (ISWC) 2017
arxiv created 2017/08/30 · openalex publication_date 2017/08/30 · arxiv updated 2017/08/31 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Nanopublications are a concept to represent Linked Data in a granular and provenance-aware manner, which has been successfully applied to a number of scientific datasets. We demonstrated in previous work how we can establish reliable and verifiable identifiers for nanopublications and sets thereof. Further adoption of these techniques, however, was probably hindered by the fact that nanopublications can lead to an explosion in the number of triples due to auxiliary information about the structure of each nanopublication and repetitive provenance and metadata. We demonstrate here that this significant overhead disappears once we take the version history of nanopublication datasets into account, calculate incremental updates, and allow users to deal with the specific subsets they need. We show that the total size and overhead of evolving scientific datasets is reduced, and typical subsets that researchers use for their analyses can be referenced and retrieved efficiently with optimized precision, persistence, and reliability.