Podcast: Play in new window | Download
Subscribe: Apple Podcasts | Spotify | Amazon Music | Android | Youtube Music | RSS
The Knowledge Graph Conference is one of the premier events in the semantic technology space.
François Scharffe and Thomas Deeley started the conference to bridge the gap between academic researchers and industry practitioners. The community they have built around the conference and the conference programming – a mix of workshops, classes, presentations, and demos – reflect this purpose.
They also intend to democratize knowledge graph use, and toward that end are participating in a number of efforts to develop education programs – both professional certifications and academic curricula.
We talked about:
- the origins of the Knowledge Graph Conference, and its balanced inclusion of both academic and industry entities
- how generative AI is propelling interest in knowledge graphs
- their mission to broaden awareness of knowledge graph technology and practice
- the importance of education and the need for a practical approach
- the origins of the Open Knowledge Network, part of the National Science Foundation’s Proto-Open Knowledge Network program, and KGC’s role in it
- their ambition to build an educational institute around KG technology
- how to help enterprise executives understand the benefits of knowledge graphs
- how modeling an enterprise’s knowledge and capturing it in a knowledge graph can help organizations address complex challenges
- the advantages of knowledge graphs over LLMs and GenAI, which have yet to prove their reliability
- how LLMs can assist in the construction and use of knowledge graphs
- how study of the human brain illustrates how GenAI and KGs can work together
- the community that has arisen around the Knowledge Graph Conference
François’ bio
François Scharffe is a hands-on technology executive with a track record of improving decision making in complex data environments. His career as a technical leader and entrepreneur has led him to perform engineering, product management and leadership roles in various organizations.
François has also worked as a lecturer and researcher, most recently at Columbia University (New York) and at the University of Montpellier (France). François is the founder and chief executive of The Data Chefs, a data management consulting firm, and the founder of the Knowledge Graph Conference, the leading event on knowledge-centric AI technologies.
Thomas’ bio
Thomas Deely is co-founder of The Knowledge Graph Conference and Community, a global community bridging research and industry on Knowledge Graphs, AI, and related technologies. Thomas is also the Customer Community Manager at Box, the leading cloud content management platform.
Thomas started his career as an engineer at JPMorgan in London, before joining Goldman Sachs where he advanced to become a senior engineer in the NY office. Thomas launched an Applied Analytics program and executive education initiatives at Columbia University, before venturing into the customer experience, product, and community domain at companies such as Unqork, where he launched the Community, and Stack Overflow, where he helped grow and develop the StackOverflow for Teams product and business, as part of the customer success team, before joining Box.
Thomas has an electronic engineering undergraduate degree from University College, Dublin, and a Masters in Science in Technology Management from Columbia University, and lives in NY where he is married with two children.
Connect with François and Thomas online
Video
Here’s the video version of our conversation:
Podcast intro transcript
This is the Knowledge Graph Insights podcast, episode number 1. I’m really happy to launch this new podcast with a conversation with François Scharffe and Thomas Deeley, the founders of The Knowledge Graph Conference. Their annual gathering in New York City attracts knowledge graph practitioners, researchers, and vendors from around the world for a full week of workshops, presentations, and tech demos. To further advance their democratization of knowledge graphs, they’re also launching new educational offerings.
Interview transcript
Larry:
Hi, everyone. Welcome to episode number one of the Knowledge Graph Insights podcast. I am really delighted today to welcome to the show François Scharffe and Thomas Deely. They are co-founders of The Knowledge Graph Conference. François is an independent consultant and an entrepreneur. He’s also got an academic position he’s on leave from right now. Thomas manages… A community manager for Box, a content sharing cloud application. So welcome, folks.
Larry:
François, do you want to tell the folks a little bit more about what you’re up to these days?
François:
Sure. Hey, Larry. Thank you for the invitation. Very glad to be here. Such an honor. First episode of what will be another eight years long-running podcast hopefully. So very cool. Thanks.
François:
Yeah, I’m up to a number of projects. I’m kind of emerging off parenting season with two small kids. And I’m looking at different opportunities always around AI and knowledge engineering, neuro-symbolic AI. And in that context, I’m looking at different topics and domains.
François:
I won’t go into the details into the projects, but I look at things like agriculture, climate. I look at education very strongly, and I think we are going to talk about this more today. And I also look at personal knowledge graphs. We published a book last year on that topic, and I think that’s an interesting space there. There’s a lot that we could do with that.
Larry:
Cool. Thanks. And Thomas, what are you up to these days?
Thomas:
Yeah. So also thanks Larry for hosting us for your first in this series. I expect it’ll be the first of many. So my day job, I work at Box, where I’m community manager. Box is a SaaS application which makes it easy for people to securely share content internally and externally. And it’s interesting times right now with AI, generative AI, and Box has some interesting propositions in that space.
Thomas:
And then my passion project is obviously KGC. And this year we were excited to get some NSF funding to build the community around knowledge graphs around education and community. So that’s also something that I’m really excited about. Yeah.
Larry:
Well, let’s talk, there’s a nice bundle of stuff to talk about there. I want to talk first about the founding of KGC because that’s kind of where all this fun stuff starts. But very quickly, after this comes the Open Knowledge Network and the educational stuff. I want to talk about that whole bundle of activities.
Larry:
So how did you… First of all, if I recall correctly, KGC came up when you were both at Columbia. Is that how it happened… Or tell me the start.
Thomas:
Yeah. So we were both working at Columbia University. And I was building out an executive education program and looking for faculty to come up with ideas. And I connected with François and he had the idea of taking, he felt there had been a lot of academic conferences on this domain, and doing something very industry focused.
Thomas:
So that was the genesis. We got some budget from Columbia to get it off the ground. And the first conference, May 2019, went really well, and that was really the genesis of KGC.
Larry:
Cool. And François, can you tell me more about that connection. Because this is something that’s always struck me about the conference, is the connection between the academy and industry. Were you the mastermind behind that part of it?
François:
Well, so mastermind is a big term. But basically for me, it was building my ideal conference, the one I missed and I would love to attend. I had worked for half of my career in academia, and there were many events in the space on that topic in academia. But after that, I devoted to mostly industry. And there were no events that were about this topic and relevant.
François:
And so being at Columbia University in New York, having tons of contacts in the field, it sounded like, well, maybe we should just start it, start that event. And then naturally, evolve from the contacts. Having contacts both in industry and in academia, we could reach and invite speakers, and there’s that flare. So over time, we’ve evolved it more towards the industry side. But indeed, we have a lot of… Our community also has a lot of academics in it.
François:
I think bridging the gap is very important. There are a bunch of things that are important and that were the goals, like the original goals. One is really… Originally the real one was say, “Hey, knowledge graphs are not an academic topic.” There’s tons of people using it for solving real problems in the industry, knowledge graphs in production. Let’s give them a voice. Let’s hear them. Let’s have a forum where we can share our experience, our issues.
François:
But also let’s bring academia to tell us about state of the art, but also to hear about what the problems actually working with this technology in the real life, what their problem are. So that can influence researchers to say, “Maybe we should give priorities to this problem in our research because they’re more important than others,” to give you an example of the kind of interactions that is making possible. Yeah. I’ll stop there.
Larry:
Thomas, did you have anything to add to that? Because I know you’re.. I perceived of you as more the business… I mean, you’re both academics, of course, but…
Thomas:
Yeah. I think this is a fascinating space. It’s arguably one of the most interesting spaces in technology. And particularly generative AI is propelling this field forward. And I think the goal was to bring an industry lens to a really interesting space, democratize it, raise awareness.
Thomas:
So our mission, we had a tagline at the conference this year is, “Where leaders create the future of knowledge together.” So our mission is really to broaden awareness of this niche topic. And why is it still niche? So make it easier for people to learn through education, and connect them with your peers and leaders in the space, people like yourself and others, so they can be somewhat inspired. And make it easy for people to dip their toe in the water, and that will ultimately drive more awareness or adoption. Because the necessity for this technology is becoming even more critical with the emergence of AI and a lot of risks that are coming with that.
Thomas:
And do it in a bottom-up community-driven way where it’s welcoming to the newcomers. We’re recognizing people who’ve been contributing to the space over years. We’re creating a platform for people who are interested to teach and who have something to say. And it’s evolved from a two-day event into a week long event, as you know yourself, over the course of the years.
Larry:
Yeah, the program itself kind of exemplifies what you talk about about democratization and education. I think the last few years anyway, it’s been two days of workshops and masterclasses, two days of programming, and then a day of demos, which is really great.
Larry:
But I’m really curious about… And so there’s a really strong, those two full days of education at the start of it. But there’s also the Open Knowledge Network. I’m not a hundred percent clear on the relationship there, that that came along, and I know the NSF has funded it – the National Science Foundation – has given it some funding. Can you talk a little bit about the OKN and maybe just education in general around knowledge graphs?
François:
So as we said, we both Thomas and I have an academic anchor at some point in our career. Back and forth for me. And so education is dear to us. So I’m a professor, I’m teaching. But there are issues with the traditional educational model. We also recognize that.
François:
So knowledge graph is a technology or a set of technologies that can be complex to grasp at first. And so since day one, our goal has always been democratizing access to this technology. So it goes through education. And while organizing a conference, as a volunteer, this is not our main job for either of us, it is requiring a lot of time and energy. And we never had the time and energy to go beyond the organization of the conference.
François:
So it was still over the years was always a topic. We tried different things but never managed to really bring something to life, until last year we got a grant from the NSF that brought some funding that is now allowing us to really become serious on that topic of education.
François:
So this is part of the National Science Foundation. The National Science Foundation has this program called the Proto-Open Knowledge Network program, which is funding a dozen of projects that are in three different themes. The main theme, which has over 10 projects, theme one, is about building knowledge graphs in different domains. So there will be education, homelessness, social justice, climate, wildlife, preservation. So a number of topics like that. But also there are supply chain, so more industrial topics as well. So each of this project is building a knowledge graph on this project. Then there are two platform projects that are building platforms to host and combine these knowledge graph into a large open knowledge network.
François:
And then there’s one theme three project, which we are part of, called Edugate, which is about education around knowledge graphs. So how to make the link between, and this is really our role in that project, to make the link between these project, which are mainly academic and government agencies led and the general public and the industry. So how to make sure that these are known first and then used. And then how to educate people around this technology, how to bring training around these technologies to the wider public. So that’s our mission.
François:
And concretely, this gives us funding for some events as part of KGC, but also developing an education platform for learning about knowledge graphs. And as well, we’re running a series of hackathons.
François:
And I think maybe Thomas can discuss this, but it creates this whole, with the conference, creates that virtuous circle, and I’m missing a couple components there, between innovation, training, and dissemination of research. But I’ll let Thomas, the architect of that model, talk about that.
Thomas:
Yeah. If there was something I learned from my years in education, both on the academic but also on the administrative side, I would say education in the US is on shaky foundations. It’s quite expensive. Often it takes a while for institutions to keep up with what’s happening in the real world. Students come out, often they can’t necessarily do anything practical. Faculty don’t get paid that much.
Thomas:
So the idea is if we can build almost like the equivalent of an educational institute around this really interesting technology domain, knowledge graphs, and related technologies, where you start… Now thanks to the NSF, we can fund a curriculum. So we’re building a curriculum right now that will make it easy for technical and non-technical people and business people and champions to have some capabilities that they can then bring into the enterprise to help drive adoption. Because one of the challenges with this space is it’s not as widely as adopted because of perceived complexity. Perhaps the benefit, the payoff takes a little longer.
Thomas:
Building into the curriculum, and this is something I did at Columbia, I ran a capstone program where before graduated, you had to complete a real world project with a real company. So François mentioned the hackathon. So before you can graduate from the KGC or to get the KGC certificate, you will have to build something aligned with some of those social-good missions that the NSF has articulated: climate, social justice, cybersecurity, supply chain.
Thomas:
And then, you may be aware, Larry, we have a startup pitch, which we’ve been running for four years as part of the conference. So perhaps now you can build something. Perhaps you want to deploy it and try and implement it. So we also have a startup pitch that we run.
Thomas:
And then the centerpiece of all that is the conference and the lifetime achievement award. And making it all the way from beginner learner to someone who’s inspired others, completing that full circle around this niche space. And we think that’s something that can help elevate and increase awareness and advance adoption.
Larry:
We’re not going to have time to complete that circle in this short conversation today, but I wonder if we could get just a quick overview. You both alluded to… I mean, this is such a powerful technology. Many obvious benefits to us anyway. But if you were just trapped in the elevator with the CEO of a company that could obviously benefit from knowledge graph technology, could one or both of you kind of make your pitch?
Larry:
And actually that’s always the wrong way to do it. You don’t pitch the technology, you find out what their problems are and help them solve it. So I guess maybe let me rephrase that and ask, what kind of business problems can a knowledge graph stack of technology help?
François:
The question is what kind of investment do you do in data management? How ready is your enterprise to use data for machine learning and AI? And the answer will be, “Oh, we bought a data governance platform,” or, “we implemented a data warehouse, we have Snowflake.” And it’s like, do you think this is going to be helpful with automated access to your data in order to build an AI, a machine learning? How ready are you for the new wave?
François:
And indeed, so especially now, people are like, “Yes, we’ve tried things where we managed to learn to build a chatbot using RAG, et cetera. It doesn’t work very well. And it took a long time because we have to gather the data and these projects are taking so long, et cetera, et cetera, et cetera.”
François:
So this is a longer term effort. Yes, you need to put effort in having your books sorted in your library. And maintain that organized, that structured way to access your knowledge like librarians. And knowledge have roots in that as well, in that domain, that library domain as well. You have catalogs, you know where the books are, they’re organized, they’re referenced. And when you need something, you have a system that is maybe a little complex and needs to be maintained in order to very quickly be able to identify where is the things you need.
François:
And we can take that analogy and take it to enterprise data right away. It takes time. It’s not just, “Hey, read the book and throw it on the pile with the rest.” No, you have to take the time and the effort to classify, to connect.
François:
And it goes, if I may, now I’m pushing it a little bit, hopefully it will still be understandable, but you also go inside the books. It’s not just the classification of the book topic, but you go inside and you link every idea in the book right together to form that connecting knowledge, which is no more in the minds of the reader, but it’s also explicit and accessible.
Larry:
I love that. The idea of just backing a dump truck up to an institution, dump a pile of books, and then you’re like, “Okay, we got to organize this, understand what’s in there.”
Thomas, do you have anything to add to that? How would you articulate the benefits and the capabilities of knowledge graph technology?
Thomas:
Yeah. And I think this will be one of the first things we’ll have in our curriculum actually. Because we’ve seen various definitions, different people have different ways of positioning it. And I think to explain it fully, you have to be able to simplify it. So there’s a few ways that… I usually start with a smarter database. So why is it smarter? Relationships are a first class citizen. Why does that matter? You get more context in your answers.
Thomas:
And then I also try and articulate it to a use case that relates to the common person. So if you’re using, let’s say Apple Siri, and how does it have context and understand what you’re asking to give you an answer that somewhat might make sense? Obviously there’s AI models there, but underneath, there’s a knowledge graph powering the infrastructure behind these tools. So making it relatable to the lay person.
Thomas:
And then in the enterprise, the use cases are many. And often they might be the more complex use cases, which are becoming more important because AI is raising the bar for what should differentiate a company in terms of projects they’re working on. And whether that’s understanding the collective intelligence of your organization by creating a social graph of all your experts across all your offices throughout the world, how powerful is that? And that’s just one of many. There’s fraud detection where you need to link often disparate data sets. You need to understand the concept of entities, and was this Larry Swanson Utrecht or Seattle?
Thomas:
And those are the areas where knowledge graphs play very well. And I think as François mentioned, it is a little bit of a pay it forward. You have to be more ambitious about the problem you’re trying to solve, and then focus on putting in the effort upfront to organize yourself to get those results.
Larry:
And as you’re talking, you’re reminding me, it seems like a good time because there’s been a lot of attention, every enterprise in the world is just awash in data these days. And there’s been data lakes and data warehouses and data meshes and all these metaphorical architectures and procedures for dealing with that.
Larry:
And it sounds like, from what you’re both saying, that knowledge graphs have that ability to organize the pile of books and to help you understand the collective intelligence of your organization. Is that unique to knowledge graphs or could you do this other ways?
Thomas:
So I-
François:
Sorry, go ahead. Go ahead, Thomas.
Thomas:
I would say probably yes. But I would say knowledge graphs are arguably one of the most powerful and effective ways and efficient ways of doing this. A great talk at our conference last year was Denny Vrandečić. He talked about the efficiency of querying these data sets using knowledge graphs versus these models which are consuming massive amounts of compute for very basic queries. So I think it’s the most powerful and arguably the most efficient way of solving these problems.
Larry:
François, you were going to say something?
François:
No, nothing that… Well, I mean, you mentioned this concept of data fabric and data mesh. I think knowledge graphs are a way to implement some of these concepts. So it’s more like a practical way to implement of these concepts. It’s more technical and pragmatic, I would say. Yeah.
Larry:
Yeah. No, I know it’s… I almost hesitate to raise this because it’s the surest way to kill these conversations around knowledge graphs. But I think the power of it is just the way they’re built. It’s like you have an ontological understanding of an organization. You know what you know, and can associate individual data things with that knowledge. That just seems more powerful than schemas that identify traditional relational databases or things like that.
Larry:
I guess without going into too much of the technical details, is it that… And maybe Thomas, maybe to ask you to follow up on that ability to understand the collective intelligence of an enterprise. That sounds like something that knowledge graphs might be uniquely suited to.
Thomas:
Yes, because you have to… It kind of requires you to model and understand your organization. Who is Larry? Where does he live? What are his interests? Where did he go to university? And spend that time to model that so you can have the context. And it’s very difficult to do that in another way, and it’s almost impossible to do it efficiently. You can do the pattern mashing of machine learning models, but I go back to the word efficiency as well here. And Denny’s talk last year where he spoke about the level of compute to get the same answer using a knowledge graph was a tiny fraction of what other methods being used.
Thomas:
So I would say there’s only certain… They come to their fore around certain types of queries and challenges which are more complex to figure out.
François:
And I think an greater advantage of knowledge graphs compared to GenAI and large language model is that they have a crisp reasoning. They do not make up data. One, when you ask a question, if you get an answer, the answer is correct. If you get an answer, because maybe you won’t get an answer if the data is not there. When you use LLM, as we know, there’s always an answer. But if there’s no answer, the answer will be there but incorrect, hallucinated. And so that’s I think a very important aspect.
François:
And we’re seeing a lot of backlash against AI where companies that have been putting consumer products, AI products, out there are backtracking because people quickly see that this is a scam, basically. It doesn’t work. It doesn’t give you good answers. It’s not reliable. And so then your brand and your products become associated to a scam, and that’s hard to get rid of. So knowledge graphs are the thing that make your system reliable. This is a reliable set of technologies.
François:
And not saying that is one or the other. The sweet spot will come from a combination of having… So that was a topic that was heavily discussed at this KGC. And what kind of emerges out of this is that there’s interaction in both directions. So now, as we mentioned several times during these discussions, knowledge graphs are technologies that are complex to put in place. They require a lot of engineering efforts.
François:
But the good news is that GenAI, large language models, are accelerating that process. They are making it easier. They automate a lot of the tasks. They allow to hide a lot of the complexity. And I can take an example, if you take RDF knowledge graph, or want to model an OWL ontology, you will have to get into those languages, have maybe somewhat of a complex syntax for the newcomer. But if now you can easily use a LLM to start to write for you and correct your errors, et cetera.
François:
And so same thing with the query language. You can use your natural language and translate automatically that query to a SPARQL or language in order to help you. And then you may revise it, but that will help you there as well. So GenAI helps building knowledge graph more efficiently.
François:
And then on the other hand, there’s a lot of research that’s using knowledge graphs in order to limit the hallucinations or improve the retrieval capabilities of large language models. So you may have heard about GraphRAG, which is combining knowledge graphs and retrieval augmented generation in order to retrieve better results and more grounded results. So grounding the results of the LLM into facts from the knowledge graphs.
So all of this is very encouraging, I think for technology in general.
Larry:
Yeah. No, it’s very exciting that so many of the problems of the LLMs and the GPTs can be helped with this technology.
Larry:
But hey, I can’t believe it. We’re coming up close to time already. But before we wrap up, I want to give each of you a chance. Is there anything last, anything you want to revisit from the conversation, or something you would just want to make sure we share before we wrap up? And either of you go.
Thomas:
Yeah. One of the reasons I find this space fascinating is a lot of the research that’s driven generative AI and these discoveries over the last 18 months or so, it comes from study of the brain. And François mentioned it’s not just one tool or the other, it’s all these tools working together. And if that’s how our brain works, we have a fuzzy part of the brain that if you saw a spider in front of you, jump back. But if you saw it from a few feet away, you wouldn’t because you have time to reason. And that’s the knowledge graph part of your brain, and then there’s the non-symbolic part of your brain. And that’s how our brain works and that’s what we’re building in these tools.
Thomas:
And there’s a great analogy of the butcher on the bus. I don’t know if you’ve read that paper. If you’re on your way home from work and someone walks on the bus. And you know the face, but you don’t know who that person is. And then on Saturday you walk into your butcher store and you see that person, “Oh, it’s my butcher…” because there’s the context around it.
Thomas:
So for me, this place is fascinating. And it’s these tools working together, with knowledge graph being the reliable reasoning part, and then there’s the fuzzy part. And having an overall view of the tools to get the best outcome, the most reliable outcome, and the most efficient outcome. And I think this is going to be going on for the next years and decades.
Larry:
Thanks. François, anything last?
François:
Last words? No, it’s certainly an exciting field to work on. We have a great community, I think. So for the first year, I was able… Because we professionalized the conference over the years. For the first year, I was able to just go to the conference and attend it. So finally the dream of building my perfect conference and being an attendee in it and sitting there and enjoying all the talks that I actually had an influence of bringing, that was really a great feeling.
François:
And also seeing that amazing community. A lot of the feedback I got was people say, “You know, the great thing at this conference that people are nice. Everyone seemed to be happy to be there and enjoying working in that field. And this is such a great community. This is such a difference from other conferences in the field I’ve been to.” So that feels good. We built a great community with a great program. And excited for next year already.
Larry:
Yeah, I am too. Well, thanks so much. Hey, and you both mentioned Denny’s keynote last year, so I’m going to include that in the show notes. And I Googled that butcher on the bus anecdote. I’ll put that in the show notes as well. I’ll ask you both in a follow-up note if there’s other stuff we might want to share. But thank you so much. This is such a great way to kick off this series.
Thomas:
Yeah, I think Juan Sequeda’s paper on benchmarking, that was another. That’s another… Yeah, so there’s probably more. But yeah, those are a few.
François:
We’ll follow up. Yeah.
Larry:
Yeah. No, I’ll definitely… And that’s something I haven’t done in my other podcast, but I want to do more in this one is just resources galore. Because there’s plenty of learning to be done, and yeah, you guys know how to help with that.
François:
Thanks, Larry. Good to talk to you.
Thomas:
Thanks, Larry.
Larry:
Thanks to both of you. Cheers.
