The Solution of WhoIsWho-IND-KDD-2024

Zhe Zhang, Weili Guo, Qing-Yuan Jiang

Keywords: author name disambiguation

Abstract:

The rapid growth of online publications has complicated the
problem of disambiguating authors with the same name, resulting
in errors in author rankings and award fraud. To tackle this, the
OGA-Challenge Team published a large-scale dataset and hosted
the KDD Cup 2024 Challenge to detect paper assignment errors
based on author and paper metadata.
In this paper, we propose a method using two strategies. The
first strategy involves extracting features from papers and authors,
and using machine learning techniques, specifically tree models,
for prediction. The second strategy constructs a graph neural network to capture the relationships between authors and papers. By
integrating these two approaches, our method achieves promising
results in detecting misassigned papers.

Submission Number: 33