Skyblocking for entity resolution

Abstract

Inthispaper,weintroduceanovelframeworkforentityresolutionblocking,calledskyblocking,whichaims to learn scheme skylines. In this skyblocking framework, each blocking scheme is mapped as apoint to a multi-dimensional scheme space where each blocking measure represents one dimension.A scheme skyline contains blocking schemes that are not dominated by any other blocking schemes intheschemespace.Toefficientlylearnschemeskylines,twochallengesexist:oneistheclassimbalanceproblem and the other is the search space problem. We tackle these two challenges by developing anactive sampling strategy and a scheme extension strategy. Based on these two strategies, we developthreeschemeskylinelearningalgorithmsforefficientlylearningschemeskylinesunderagivennumberof blocking measures and within a label budget limit. We experimentally verify that our algorithmsoutperform the baseline approaches in all of the following aspects: label efficiency, blocking qualityand learning efficiency, over five real-world datasets. Show more