Alan Morrison of GraphRAG Curator recently interviewed me about my work both inside and outside of Graphwise, and it was a lot of fun. You can see it on YouTube and read a transcript on the GraphRAG Curator website .

Reading the transcript later, it felt like a first draft that I’m not allowed to edit. So, for example, when Alan asked “if there was some really common issue across companies that you knew you could change, what would you change to help us out with the governance problem that we have with data?” my answer was “It might be a cop-out, but for any given company, what are you trying to do?” In other words, I couldn’t think of general advice and would find it better to focus on each company’s individual goals.

Later, though, I thought of something that could potentially help all companies, at least from a knowledge graph perspective: doing a complete inventory of all of your organization’s information assets might turn up some nice surprises for your enterprise knowledge graph. The first few things that you find will be your main systems that you use most often, but keep looking and you may find things you forgot about that relate to the first few datasets you found and can enrich your use of them. A little extra metadata for each dataset may be all you need to get them making a contribution to your knowledge hub.

For example, let’s say a few years ago someone did a special project about a certain subset of their company’s inventory and the result was a set of spreadsheets. People might be thinking “well, that wasn’t official company data but just some back-of-the-napkin experiments and would therefore not be in the domain of serious data governance”, but maybe it could still contribute in some interesting new ways. For example, after converting the spreadsheets to RDF, which is easy enough, the insights gained from that project could become a long-term part of the more official data about those inventory items, even though these old spreadsheets would typically not be considered one of the company’s important assets. The flexibility of RDF makes it simple to add a few new properties to a few entities in your dataset with no need for some schema revision procedure.

Here are links to some of the projects that Alan and I discussed, ranging from serious videos about Graphwise GraphDB capabilities to fun things I’ve done on my own, but nearly all with some RDF angle:

• Recent blog entry Let’s stop saying “semantic web”

• 2021 blog entry You probably don’t need OWL

• The video Build a shopping chatbot in four minutes with GraphDB Talk to Your Graph 2.0

• The video Using GraphDB’s Lucene Connector

• My 2017 blog entry SPARQL queries of Beatles recording sessions, which includes a link to the RDF data that you can download

• My 2016 blog entry Converting between MIDI and RDF: readable MIDI and more fun with RDF

• 2012 blog entry A brief, opinionated history of XML

• IPTC’s specialization of schema.org for sports data

• Bandcamp’s Best contemporary classical roundups

• Is the Dewey Decimal System available as a SKOS dataset? The W3C’s SKOS/Datasets points to one at http://dewey.info, along with a SPARQL endpoint, but neither of those links seem to work anymore.

I hope the interview is interesting if you are interested in knowledge graphs and Graph RAG applications.

Comments? Reply to my Mastodon or Bluesky posts announcing this blog entry.