Best paper award at IEEE DSAA 2023
AAII's Jiaxi Li receives Best Paper award for 'Heterogeneous Graph-level Anomaly Detection' at the 2023 IEEE International Conference on Data Science and Advanced Analytics (DSAA).
State-of-the-art anomaly detection approach awarded 'Best Paper' at IEEE DSAA 2023
AAII's Jiaxi Li and Ling Chen have been awarded 'Best Paper' (Applications Track) at the recent IEEE International Conference on Data Science and Advanced Analytics (DSAA) which took place in Thessaloniki, Greece, October 9 to 12 2023.
DSAA'2023 provides a premier forum that brings together researchers, industry and government practitioners, as well as developers and users of big data solutions. The conference is widely recognised as a flagship annual meeting in data science and analytics.
The awarded paper, 'Heterogeneous Graph-level Anomaly Detection with Hierarchical Relation-augmented Graph Neural Networks', considers the problem of heterogeneous graph-level anomaly detection.
It proposes HRGCN: an unsupervised deep heterogeneous graph neural network, to model complex heterogeneous relations between different entities in the system for effectively identifying these anomalous behaviour graphs.
HRGCN trains a hierarchical relation-augmented Heterogeneous Graph Neural Network (HetGNN), which learns better graph representations by modelling the interactions among all the system entities and considering both source-to-destination entity (node) types and their relation (edge) types.
DSAA'2023 Application Track
The DSAA'2023 'Application Track' attracts innovative research and best practices of Data Science and Advanced Analytics across various disciplines and domains including business, government, health and medical science, physical sciences, and social sciences, with expected practical applications on real-life data.
The innovative HRGCN anomaly detection approach awarded 'Best Paper' in this category has demonstrated practical use, with extensive evaluation on two real-world application datasets showing that HRGCN outperforms state-of-the-art competing anomaly detection approaches.
Within the paper, researchers also present a real-world industrial case study to justify the effectiveness of HRGCN in detecting anomalous (e.g. congested) network devices in a mobile communication service.
Detecting anomalous heterogeneous graphs from a large set of system behaviour graphs is crucial for many real-world applications, such as online web/mobile service, cloud access control and capturing information in complex industrial operating systems.
HRGCN is available at https://github.com/jiaxililearn/HRGCN.