Exploiting Source-Object Network to Resolve Object Conflicts in Linked Data

TL;DR

This paper introduces a novel network-based method called ObResolution for resolving conflicting objects in Linked Data by modeling correlations in a Source-Object Network, demonstrating superior accuracy and robustness across multiple datasets.

Contribution

It formalizes the object conflicts resolution as a joint distribution problem on a Source-Object Network and proposes a pMRF-based approach for accurate conflict resolution.

Findings

Higher accuracy than existing methods

Robustness across various domains

Effective modeling of correlations in data

Abstract

Considerable effort has been made to increase the scale of Linked Data. However, an inevitable problem when dealing with data integration from multiple sources is that multiple different sources often provide conflicting objects for a certain predicate of the same real-world entity, so-called object conflicts problem. Currently, the object conflicts problem has not received sufficient attention in the Linked Data community. In this paper, we first formalize the object conflicts resolution problem as computing the joint distribution of variables on a heterogeneous information network called the Source-Object Network, which successfully captures the all correlations from objects and Linked Data sources. Then, we introduce a novel approach based on network effects called ObResolution(Object Resolution), to identify a true object from multiple conflicting objects. ObResolution adopts a pairwise Markov Random Field (pMRF) to model all evidences under a unified framework. Extensive experimental results on six real-world datasets show that our method exhibits higher accuracy than existing approaches and it is robust and consistent in various domains. \keywords{Linked Data, Object Conflicts, Linked Data Quality, Truth Discovery