2023年8月30日，我院张帆教授受邀线上参加韩国蔚山国立科学技术院（UNIST）CSE×AI研讨会，并就图数据计算研究进展做1小时报告。报告介绍了课题组团队在图计算领域多个方向的研究成果，包括稠密子图、大图分解、图稳定性等模型，以及面向大规模图数据的高效静态算法和动态更新技术。此外，报告也涵盖异常检测、信息传播等方向的研究成果，以及关键计算模块的未来研究方向。报告结束后，双方老师同学们对报告内容进行了积极的交流探讨。

报告题目：Efficient Graph Computation: Cohesive Subgraph, Graph Decomposition, and More

摘要：Graph can efficiently analyze complex relationships between entities and model various large-scale networks such as social networks, the Internet, and financial transaction networks. The talk will introduce the models on cohesive subgraphs, graph decomposition, and network stability, as well as efficient static algorithms and dynamic updating techniques for processing large-scale graph data. Cohesive subgraphs are formed by a group of tightly connected vertices and serve as fundamental components of graph data. The computation of cohesive subgraphs can decompose the graphs and abstract the connectivity structure between vertices s.t. hidden information in the graphs can be efficiently discovered. Graph decomposition can also be used to analyze the structural stability of different networks and identify important nodes/connections that can enhance/maintain network stability. The presentation will also discuss future research directions.