Let’s Get Connected with KDD.SG Seminar on Graphs

In this seminar by KDD.SG held on April 26, 2023, we featured three experts on graphs.

Talk #1: Title: Efficient Graph Representations for Dynamic Processing on GPUs: Leveraging Dynamic Data Structures for Adaptability and Versatility

Abstract: Dynamic graph processing has become increasingly important in various fields, such as social network analysis, disease surveillance, and transaction monitoring, where graphs evolve over time or need to be analyzed in real-time. Graphics Processing Units (GPUs) offer massive parallelism and high computational power, making them an ideal choice for processing large-scale dynamic graphs. However, efficient processing of these dynamic graphs on GPUs demands adaptive and versatile graph representations. In this talk, we delve into the design of innovative graph representations that enable efficient dynamic graph processing on GPUs. We will discuss the characteristics of our graph representation, which utilizes dynamic data structures to cater to different update requirements. By highlighting the importance of adaptability and versatility in graph representations, attendees will gain valuable insights into optimizing dynamic graph processing on GPUs and be better equipped to enhance their own graph-based applications.

Bio: Yuchen Li is an assistant professor with the School of Computing and Information Systems, Singapore Management University (SMU). He received double BSc degrees in applied math and computer science (both with first-class honors) and a Ph.D. degree in computer science from the National University of Singapore (NUS), in 2013 and 2016, respectively.

Talk #2: Dense subgraph mining: problems, algorithms and applications

Abstract: A graph represents a set entities as vertices and the relations among the entities as edges and is used for modelling data in many applications. A subgraph of vertices in a graph, which are connected via many edges, is called a dense subgraph and often conveys meaningful knowledge of the graph. Finding dense subgraphs in a graph has a wide range of applications including correlation mining, fraud detection, e-commerce, bioinformatics, frequent pattern mining, and community detection, etc. In this talk, I will first introduce some representative applications, density measures and problems of dense subgraph mining. I will then present several of our recent studies of finding dense subgraphs in bipartite graphs and uncertain graphs and their applications in fraud detection, community search, and fake news detection. Finally, I will discuss some potential future work directions.

Bio: LONG Cheng is currently an Assistant Professor at the School of Computer Science and Engineering (SCSE), Nanyang Technological University (NTU). He got the PhD degree from Hong Kong University of Science and Technology (HKUST) in 2015. His research interests are broadly in data management and data mining.

Talk #3 Low-Resource Learning on Graphs
Abstract: Graph structures are ubiquitous in various domains, ranging from social networks and e-commerce platforms to transportation and biological systems. On these graphs, various graph-based analytics and mining tasks exist, many of which can be cast as instances of link prediction, node classification, and graph classification. Moving away from manual feature engineering, graph neural networks (GNN) have witnessed widespread success in various application scenarios due to their ability to learn powerful graph representations automatically. However, their success is often dependent on the availability and quality of graph structures and labeled data, without which their performance can suffer. In this talk, we explore alternative learning paradigms different from the traditional supervised learning paradigm, specifically addressing two types of low-resource scenarios on graphs: structure scarcity and label scarcity. We will first provide an overview of low-resource problems and methods on graphs, and then introduce some of our representative works on these problems.

Bio: Dr. Yuan Fang is an Assistant Professor at the School of Computing and Information Systems at Singapore Management University (SMU). He was previously a data scientist at DBS Bank and a research scientist at A*STAR. His research interests revolve around graph-based learning and its applications in recommender systems, social network analysis, and bioinformatics.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s