New measures for evaluating creativity in scientific publications

TL;DR

This paper introduces new citation-based measures to evaluate the creativity of scientific publications by analyzing citation patterns, categories, and distribution across scientific groups in high-frequency analog circuit design.

Contribution

It proposes a novel categorization of citations and two new measures to assess publication creativity based on citation patterns and group interactions.

Findings

Citation distribution follows a Levy type distribution.

Different scientific groups cite publications in distinct motifs.

New measures effectively characterize creativity in scientific work.

Abstract

The goal of our research is to understand how ideas propagate, combine and are created in large social networks. In this work, we look at a sample of relevant scientific publications in the area of high-frequency analog circuit design and their citation distribution. A novel aspect of our work is the way in which we categorize citations based on the reason and place of it in a publication. We created seven citation categories from general domain references, references to specific methods used in the same domain problem, references to an analysis method, references for experimental comparison and so on. This added information allows us to define two new measures to characterize the creativity (novelty and usefulness) of a publication based on its pattern of citations clustered by reason, place and citing scientific group. We analyzed 30 publications in relevant journals since 2000 and their about 300 citations, all in the area of high-frequency analog circuit design. We observed that the number of citations a publication receives from different scientific groups matches a Levy type distribution: with a large number of groups citing a publication relatively few times, and a very small number of groups citing a publication a large number of times. We looked at the motifs a publication is cited differently by different scientific groups.