OAG-Challenge at KDD CUP 2024 Launched!

[zfj...@gmail.com]

Hi all,

We are thrilled to announce the Open Academic Graph (OAG) Challenge at the KDD Cup 2024, aimed at addressing the pivotal challenges in academic graph mining. Participants can choose from three intriguing tasks: author name disambiguation, academic question answering, and paper source tracing. We encourage the use of large language models, and to support your efforts, we're offering free GLM API token sponsorship!

In addition, we are introducing OAG-Bench, a comprehensive, fine-grained, human-curated benchmark based on the Open Academic Graph. This benchmark is designed to encompass the entire life cycle of academic graph mining. OAG-Bench includes 10 tasks, 20 datasets, over 70 baselines, and more than 120 experimental results to date. We've developed new data annotation strategies for certain tasks and provide a suite of data pre-processing codes, algorithm implementations, and standardized evaluation protocols to facilitate academic graph mining. See more detailed information below.

Challenge website: https://www.biendata.xyz/kdd2024/
Paper link: https://arxiv.org/pdf/2402.15810
Data link: https://www.aminer.cn/data/

Huge thanks to the team members of the KEG Group at Tsinghua University, Zhipu AI, and the steering committee members for their invaluable contributions to this initiative.

--
Best,
Fanjin Zhang
Postdoctoral Researcher
Dept. of Computer Science & Technology
Tsinghua University