Domain expertise extraction for finding rising stars

Lin Zhu, Junjie Zhang, Scott W. Cunningham

Research output: Contribution to journalArticlepeer-review

16 Downloads (Pure)

Abstract

The field of expertise extraction utilizes published research enabling communities to highlight and identify the skills of researchers within specific scientific domains. This can be useful for evaluating research performance, and in the case of rising stars, in identifying top scientific talent. Previous research has harvested a range of publication indicators in an effort to identify expertise and talent. These include content indicators, citation metrics, and also the position of a researcher within a full collaboration network of scientists. The existing mechanism of expertise extraction utilizes all papers attributed to a scientific author, thereby potentially neglecting their specific or specialized expertise. Here we show that a tensor decomposition technique when applied to the problem addresses a number of useful problems. This includes better identification of individual expertise, as well as an integrated appraisal of an author’s role in an extended scientific network. The technique will afford new analyses of knowledge production which consider specialisation and diversity as core elements for further analysis. More generally the tensor decomposition techniques presented in this paper can be applied to a range of scientometric problems where multi-modal data is encountered.
Original languageEnglish
Pages (from-to)5475-5495
Number of pages21
JournalScientometrics
Volume127
Issue number9
Early online date22 Aug 2022
DOIs
Publication statusPublished - 1 Sept 2022

Keywords

  • research expertise
  • scientific indicators
  • scientometric tensor

Fingerprint

Dive into the research topics of 'Domain expertise extraction for finding rising stars'. Together they form a unique fingerprint.

Cite this