English

A Bayesian Approach for De-duplication in the Presence of Relational Data

Methodology 2021-11-17 v4

Abstract

In this paper, we study the impact of combining profile and network data in a de-duplication setting. We also assess the influence of a range of prior distributions on the linkage structure. Furthermore, we explore stochastic gradient Hamiltonian Monte Carlo methods as a faster alternative to obtain samples from the posterior distribution for network parameters. Our methodology is evaluated using the RLdata500 data, which is a popular dataset in the record linkage literature.

Keywords

Cite

@article{arxiv.1909.06519,
  title  = {A Bayesian Approach for De-duplication in the Presence of Relational Data},
  author = {Juan Sosa and Abel Rodriguez},
  journal= {arXiv preprint arXiv:1909.06519},
  year   = {2021}
}