OpenKG Academic Salon on Symbolic Knowledge and Large Models – April 12, 2024, Suzhou University

Time: April 12, 2024, 10:00‑11:50; Location: Suzhou University, Building 504.

OpenKG will host an academic salon titled “Symbolic Knowledge and Large Models” with three TOC experts from Beijing University of Posts and Telecommunications, Tongji University, and Tianjin University discussing frontier topics in knowledge graphs and large language models.

Schedule: 10:00‑10:10 OpenKG introduction (Chen Huajun); 10:10‑10:40 Graph Foundation Models (Shi Chuan); 10:40‑11:10 Knowledge‑Enhanced Large Models (Wang Haofen); 11:10‑11:40 Co‑evolution of Large Models and Knowledge Graphs (Wang Xin); 11:40‑11:50 OpenKG summary (Qi Guilin).

Report 1 – Graph Foundation Models: The rise of graph neural networks and their synergy with large language models such as ChatGPT are reshaping AI research; this talk introduces the concept of graph foundation models, related work, and preliminary explorations.

Report 2 – Knowledge‑Enhanced Large Models: Large language models achieve human‑level performance but face challenges like hallucinations, outdated knowledge, and privacy; knowledge‑enhancement techniques can address these issues and enable deep integration of AI in vertical domains.

Report 3 – Co‑evolution of Large Models and Knowledge Graphs: Large models represent connectionist AI, while knowledge graphs embody symbolic AI; the talk explores recent progress in their collaboration, proposing a neural‑plus‑symbolic fusion pathway.

Speaker biographies highlight the extensive research achievements, patents, and leadership roles of each expert in AI, knowledge graphs, and related fields.

The event will be livestreamed via the OpenKG video channel; a QR code is provided for online reservation.

OpenKG is a nonprofit, neutral, academic community dedicated to opening Chinese‑language knowledge‑graph data, tools, algorithms, and platforms, and to fostering deep integration with large models.