Jelly RDF support in LDES #94
Description
Jelly is a binary RDF format that supports streaming parsing of “frames” of quads (i.e. groups of quads that we need for members).
A quick experiment I did shows it is way faster that the current fastest LDES-based parsing method using the TREE profile
A yet to be published paper presented at SEMANTiCS2025 applies it on Nanopublications in a set-up that is similar to LDES with promising results.
Combining the jelly serialization format with LDES’ retention policies, fragmentations, and other context information might yield interesting results.
In order to get Jelly support for LDES, we should define the order of the quads in this binary format, and indicate how to make use of the frames. I would propose to use a similar system as the TREE profile algorithm.