Abstract:Cryptocurrencies like Bitcoin have provided a platform for parties to transfer digital value in a decentralised way. However, to support scalability, the amount of extra data that can be attached to a transaction is very limited, greatly limiting the possibility of attaching descriptive metadata to them with the same guarantees that the Blockchain
In this paper we study the problem of attaching metadata in RDF to Bitcoin-style transactions, where extra data is limited to 83 bytes per transaction, linking the problem to the one of RDF-Compression. However, the motivating scenarios for RDF compression (IOT devices and Big Data Publish/Exchange) do not have the same requirements as ours, as memory is extremely limited, but the input data is small and there is no need for energy efficiency or providing a query interface. We evaluated the size reduction of 7 RDF compression algorithms on two small documents (4 and 10 triples), finding that only two of them on 4 triples and none on 10 triples were able to improve over a naive gzip compression. Our findings motivate the need of further research on adequate representations for this particular use case.
TL;DR:Tested SoA RDF compression algorithms for the use case of attaching data to Crytptocurrency transactions. They don't perform well.