Better Retrieval, Limited Clustering Gains: A Controlled Study of Multilingual Company Entity Resolution
Improved name retrieval may have little effect on company clusters when the pair classifier remains unchanged. We examine this dependency by adapting multilingual E5 encoders under fixed candidate budgets and downstream decision rules. Random-negative and hard-negative training use identical positive schedules. Checkpo...
Yijiashun Qi, Yuxuan Li, Han Guo
· 0 citations