Estimating the Performance of Entity Resolution Algorithms: Lessons Learned Through PatentsView.org

TL;DR

This paper presents a new evaluation method for entity resolution algorithms, specifically applied to PatentsView.org, providing a practical way to assess and compare disambiguation performance considering sampling biases.

Contribution

It introduces a data collection and performance estimation approach tailored for patent data, enabling reliable assessment of disambiguation algorithms.

Findings

First representative performance evaluation of PatentsView disambiguation

Method accounts for sampling biases in patent data

Facilitates comparison of different entity resolution algorithms

Abstract

This paper introduces a novel evaluation methodology for entity resolution algorithms. It is motivated by PatentsView.org, a U.S. Patents and Trademarks Office patent data exploration tool that disambiguates patent inventors using an entity resolution algorithm. We provide a data collection methodology and tailored performance estimators that account for sampling biases. Our approach is simple, practical and principled -- key characteristics that allow us to paint the first representative picture of PatentsView's disambiguation performance. This approach is used to inform PatentsView's users of the reliability of the data and to allow the comparison of competing disambiguation algorithms.