Ph.D. student Wan Qizhi from the School of Information Management has published a paper in an internationally renowned journal

Recently, Ph.D. student Wan Qizhi from the School of Information Management has published a paper as the first author in the internationally renowned journal ACM Transactions on Information Systems (TOIS) in the field of data mining. The paper is titled "A Multi-Channel Hierarchical Graph Attention Network for Open Event Extraction". It is the first paper published by our province in this journal and represents another significant advancement made by our university in the field of data mining.

TOIS is one of the most influential international journals in the field of data mining. It is recommended by the China Computer Federation as an international Class A journal in the areas of databases, data mining, and information retrieval, and it is recognized as an A-class international academic journal by prestigious institutions such as Tsinghua University. Published by the Association for Computing Machinery (ACM), TOIS releases four issues per year, featuring approximately 40 outstanding academic papers annually.

        The paper addresses the problem of open event extraction. It constructs a bidirectional dependency graph based on syntactic dependency trees, which contains rich semantic and structural information. The paper improves the original graph attention model and proposes a multi-channel hierarchical graph attention network. This approach not only enables the extraction of open events, breaking the limitation of requiring specified event types, but also solves the problem of unknown event types in many application scenarios. Additionally, it provides a feasible method for understanding the core semantics of documents and sentences, assisting users in quickly grasping the content expressed in the documents.