PromptEM

Abstract

Entity Matching (EM), which aims to identify whether two entity records from two relational tables refer to the same real-world entity, is one of the fundamental problems in data management. Traditional EM assumes that two tables are homogeneous with the aligned schema, while it is common that entity records of different formats (e.g., relational, semi-structured, or textual types) involve in practical scenarios. It is not practical to unify their schemas due to the different formats. To support EM on format-different entity records, Generalized Entity Matching (GEM) has been proposed and gained much attention recently. To do GEM, existing methods typically perform in a supervised learning way, which relies on a large amount of high-quality labeled examples. However, the labeling process is extremely labor-intensive, and frustrates the use of GEM. Low-resource GEM, i.e., GEM that only requires a small number of labeled examples, becomes an urgent need. To this end, this paper, for the first time, focuses on the low-resource GEM and proposes a novel low-resource GEM method, termed as PromptEM. PromptEM has addressed three challenging issues (i.e., designing GEM-specific prompt-tuning, improving pseudo-labels quality, and running efficient self-training) in low-resource GEM. Extensive experimental results on eight real benchmarks demonstrate the superiority of PromptEM in terms of effectiveness and efficiency.

PromptER: Prompt Contrastive Learning for Generalized Entity Resolution

“…We study the scheduling and coordination of the individual ER algorithms, in order to resolve the multiple datasets, and show the scalability of our approach….." The conflict between efficiency and effectiveness: The entity resolution models based on PLMs can be divided into two categories regarding representation learning: independent or interdependent representation [9]. Interdependent representation models [2,3,10,11] have a deep interaction between pairs of records through attention mechanisms, resulting in better matching quality. Despite being effective, interdependent representation models come with a poor scalability for the quadratic searching space of record pairs, thus need additional blocking steps.…”

“…By constructing prompt templates, downstream tasks can be transformed into fill-in-the-blank forms of upstream tasks, which can more effectively utilize the original network structure of PLM and the prior knowledge obtained from pretraining. PromptEM [10] is the first work that applies prompt tuning for entity resolution tasks and performs well under low-resource and sufficient resource settings. Recently, some work also adopted the prompt-based method to eliminate embedding bias in PLMs [24], which can also be used to improve the embedding quality of records in entity resolution tasks.…”