Knowledge Graph Conference 2023

In a few short years, leaders in knowledge management, data architecture, and information technology in their organizations have gone from asking questions such as “What is a knowledge graph?” and “Do we need a knowledge graph?” to “How can I most successfully implement a knowledge graph?” and “What additional applications can I build on top of my knowledge graph?”. Similarly, our conference presentations have moved on from “The Knowledge Graph—The New Unknown Beast” and “A Brief Introduction to Knowledge Graphs” in 2019 to presentations of specific applications of knowledge graphs. 

A knowledge graph is a semantically linked layer of extracted/virtual data which is between the original data repositories and frontend applications The links and semantic enrichment are based on an ontology. Because a knowledge graph has various components and layers (extracted instance data, metadata, a graph database to manage the data, a terminology/controlled vocabulary/taxonomy, an ontology, a semantic search application, and perhaps also a data catalog and a search index), building a knowledge graph requires involvement of stakeholders in different roles and areas of expertise. As a result, these different stakeholders have different perspectives and hold slightly different definitions of knowledge graphs. The differences are based on focus and emphasis, and are not in disagreement.

KGC is a completely independent conference run by a small number of volunteers and staff who are dedicated to the field. It has grown from a 2-day in-person only conference to a 4-day in-person + hybrid (live streaming and recorded) conference plus a fifth virtual-only day. The first conference was held at Columbia University in 2019, which was followed by two years of virtual conferences. In May 2022 and this year, KGC took place at the conference center of Cornell Tech Campus on Roosevelt Island, a pleasant location in the East River alongside Manhattan. KGC is more than just a conference. It is also a community of professionals from industry, academia, tools vendors, and consultants, who are interested in discussing knowledge graphs and related technologies. KGC has a Slack space, open to all, which has discussion year-round, not just around the time of the conference. There is also a related Q&A discourse platform. Furthermore, KGC hosts occasional online events such as virtual (Zoom) book club series, which in fall 2021 featured our SWC authors of The Knowledge Graph Cookbook, Andreas Blumauer and Helmut Nagy. Another book club on an ontology book has led to a spinoff book club project weekly Zoom meeting group. In the past there have been “Knowledge Espresso” hosted interviews and Q&A sessions with experts.

On the two regular conference days there were three simultaneous in-person sessions, following the daily keynotes, and two more simultaneous virtual sessions. There were themed tracks: Business Use Cases; Content Knowledge Graphs; Data Architecture; Deep Learning for and with Knowledge Graphs; Environmental, Social and Governance; Metadata; Natural Language Processing (NLP); Ontologies, Taxonomies, Data Modeling; and Semantic Layer. The keynotes and much discussion was centered around large language models (LLMs) and generative AI. 

All sessions were live-streamed and recorded on the Airmeet virtual conference platform, which are available to attendees (depending on their registration type) for three months after the conference, so I have enjoyed the opportunity to listen to talks that I missed due to simultaneous sessions. A sufficient number of dedicated volunteers helps make the logistics of a fully hybrid conference possible.

Despite the addition of the virtual option, physical attendance has increased over the three years of live conferences to over 280 attendees this year, or over 200 on each main conference day, and a much larger number online. Attendees were from all different industries and some from government agencies. They came from across the United States (as there is no other knowledge graph conference in the country) and some from Europe. The speakers and sponsors were particularly international, with many coming from Europe. 

SWC CFO and co-founder Martin Kaltenböck presented “How Knowledge Graphs foster Interoperability in Data Spaces.” This is related to Semantic Web Company’s recent award to participate in a funded research project of the Digital Europe program UNDERPIN, a pan-European data space for holistic asset management in critical manufacturing industries. 

Martin also presented a talk which was developed by SWC CEO Andreas Blumauer, who could not attend, “Supporting ESG Strategies with Enterprise Recommender Systems and Knowledge Graphs.” Andreas subsequently expanded on this subject in a PoolParty webinar, “ESG Strategies with Semantic AI.”  Andreas also presented a short talk and demo, “PoolParty for SharePoint,”  in the semantic search-themes session on the final day of pre-recorded tools talks. (More information is in the upcoming “PoolParty for SharePoint” webinar.)

We were pleased with the engaged Q&A and comments from the audiences during our presentations. 

We also had a full size exhibit booth at KGC for the first time and appreciated the opportunity to discuss the interests in knowledge graphs with booth visitors give quick demos of PoolParty and its applications, including the newly developed demo “PoolParty meets ChatGPT” which makes use of our ESG core taxonomy in PoolParty to enrich the answers from ChatGPT to ESG questions. 

We look forward to participating again in KGC 2024.