Highlights of the 2022 World AI Conference Graph Intelligence Forum in Shanghai

On September 1, 2022, the first forum of the World AI Conference, titled “Next‑Generation Graph Intelligence Technology Development and Practice Forum,” was held at the Shanghai Expo Center.

The sub‑forum brought together top global graph‑intelligence researchers such as Chinese Academy of Engineering academicians Chen Chun and Zheng Weimin, and announced the release of the world’s fastest open‑source graph database.

Graph intelligence, encompassing graph computing, graph learning, and graph databases, is regarded as the next frontier in artificial intelligence, poised to shape technological competition over the next 5‑10 years, with major tech companies strategically investing in R&D and labs.

Academician Chen Chun emphasized the importance of integrating time with graph structures, arguing that temporal graphs enable real‑time computation and intelligent decision‑making for problems ranging from trust relationships to bridge quality monitoring.

Professor Zheng Weimin of Tsinghua University presented a series of large‑scale graph computing systems and technologies developed by his team, highlighting the burgeoning opportunities in graph computing.

Professor Jin Hai from Huazhong University of Science and Technology stressed that the explosive growth of data makes efficient graph processing essential, and introduced the world’s first data‑flow‑based graph computing accelerator supporting multiple algorithms.

The “China Graph Intelligence Team” is emerging as a strong force, achieving top results in tests and gaining global recognition for its technical concepts.

UCLA professor Harry Xu described how large‑scale graph computing systems can be applied to program analysis, abstracting code into graph problems to uncover hidden vulnerabilities and address combinatorial explosion challenges.

Gabor Szarnya, head of the LDBC benchmark workgroup, highlighted the need for standardized benchmarks to accelerate graph technology development and improve system performance and maturity.

Zhang Weimin, secretary‑general of the New‑Generation AI Industry Innovation Alliance, advocated for building open‑source AI communities and sharing graph computing technologies to foster collaborative innovation.

Scholars from Shanghai Jiao Tong University and Huazhong University, together with experts from Ant Group, ByteDance, Yueshu Technology, and Amazon, participated in a round‑table discussion on open‑source models, graph‑relational data models, and the technical and commercial challenges of graph applications, emphasizing stronger industry‑academia collaboration.

Ant Group’s graph computing leader Chen Wenguang announced the open‑source release of Ant Group’s high‑performance graph database TuGraph (single‑node edition), touted as the world’s fastest graph database.

According to Ant Group’s graph database head Hong Chuntao, TuGraph, co‑developed with Tsinghua University, won the leading technology achievement award at the 2021 World Internet Conference and broke records in the LDBC SNB benchmark in August, securing the top global position.

This sub‑forum was guided by the World AI Conference committee and co‑organized by the New‑Generation AI Industry Innovation Alliance and Ant Technology Research Institute, marking the second consecutive year the conference hosts a graph‑technology forum.