An alternate approach to Lucene + fuseki, though decidedly less Open Source, is to use the MarkLogic connector for Jena. MarkLogic is a document store with highly sophisticated text analytics and search, plus it also supports triples, graphs + SPARQL. Documents can embed both complex markup, free text and triples.
Marklogic-jena is a library to expose MarkLogic to Java as a DatasetGraph. https://github.com/marklogic/marklogic-jena I understand that a commercial product like this is not always what folks are searching for, but it sure does perform well for what you want, and it scales to far vaster realms than any multi-backend system will. If it's your company looking for a solution, you might want to take a look, and I'm always lurking here if you have questions. Charles Greer Lead Engineer MarkLogic Corporation cgr...@marklogic.com Cell: +1 707 318 0442 @grechaw www.marklogic.com ________________________________________ From: Chris Wood [c.c.w...@gmail.com] Sent: Thursday, August 24, 2017 10:16 AM To: users@jena.apache.org Subject: Using Lucene with multiple triplestores A couple of years ago, there was a discussion on this mailing list about some problems that had been encountered when Lucene was installed alongside multiple triplestores [1]. Some suggestions were made as to how to resolve this issue. Were any of these ever implemented? If not, is the best solution still to have a Lucene instance per triplestore, or is there now a way of configuring Fuseki conf files to share a single Lucene instance? Best, Chris 1. http://users.jena.apache.narkive.com/EFr0CiFe/fuseki-2-services-sharing-a-dataset-with-text-index
