Estimating Thematic Similarity of Scholarly Papers with Their Resistance Distance in an Electric Network Model

TL;DR

This paper introduces a novel method for measuring thematic similarity between scholarly papers using resistance distances in an electric network model, validated on a citation network of 492 papers.

Contribution

It proposes a new approach to quantify thematic distance based on electric resistance in a citation network, offering a realistic measure validated with empirical data.

Findings

Resistance distance correlates with thematic similarity.

The method is validated on a nearly bipartite citation network.

Provides a new quantitative tool for scholarly paper analysis.

Abstract

We calculate resistance distances between papers in a nearly bipartite citation network of 492 papers and the sources cited by them. We validate that this is a realistic measure of thematic distance if each citation link has an electric resistance equal to the geometric mean of the number of the paper's references and the citation number of the cited source.