APrompt4EM: Augmented Prompt Tuning for Generalized Entity Matching
APrompt4EM: Augmented Prompt Tuning for Generalized Entity Matching
Generalized Entity Matching (GEM), which aims at judging whether two records represented in different formats refer to the same real-world entity, is an essential task in data management. The prompt tuning paradigm for pre-trained language models (PLMs), including the recent PromptEM model, effectively addresses the challenges of low-resource GEM in …