Machine Learning methods for Author Name Disambiguation
Keywords: Name Disambiguation, Embedding, Gradient Boosting Decision Tree
TL;DR: •Computing methodologies ~ Machine learning ~ Machine learning approaches
Abstract: This paper describes the Rank 8 solution of KDD CUP 2024 OAG-Challenge WhoIsWho-IND Task. The task is to develop a model to discover paper assignment errors for given authors. We take use of 3 kinds of embedding methods combining with manual feature engineering. Then we build single-models based on LightGBM and Xgboost with several subsets of features and apply an ensemble for these models aiming at a high weighted AUC.
Submission Number: 18
Loading