Active Blocking Scheme Learning for Entity Resolution

Blocking is an important part of entity resolution. It aims to improve time efficiency by grouping potentially matched records into the same block. In the past, both supervised and unsupervised approaches have been proposed. Nonetheless, existing approaches have some limitations: either a large amount of labels are required or blocking quality is hard to be guaranteed. To address these issues, we propose a blocking scheme learning approach based on active learning techniques. With a limited label budget, our approach can learn a blocking scheme to generate high quality blocks. Two strategies called active sampling and active branching are proposed to select samples and generate blocking schemes efficiently. We experimentally verify that our approach outperforms several baseline approaches over four real-world datasets.