Evaluating technological emergence using text analytics: Two case technologies and three approaches

Samira Ranaei, Arho Suominen (Corresponding Author), Alan Porter, Stephen Carley

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Scientometric methods have long been used to identify technological trajectories, but we have seldom seen reproducible methods that allow for the identification of a technological emergence in a set of documents. This study evaluates the use of three different reproducible approaches for identifying the emergence of technological novelties in scientific publications. The selected approaches are term counting technique, the emergence score (EScore) and Latent Dirichlet Allocation (LDA). We found that the methods provide somewhat distinct perspectives on technological. The term count based method identifies detailed emergence patterns. EScore is a complex bibliometric indicator that provides a holistic view of emergence by considering several parameters, namely term frequency, size, and origin of the research community. LDA traces emergence at the thematic level and provides insights on the linkages between emerging research topics. The results suggest that term counting produces results practical for operational purposes, while LDA offers insight at a strategic level.

Original languageEnglish
JournalScientometrics
DOIs
Publication statusE-pub ahead of print - 2 Nov 2019
MoE publication typeA1 Journal article-refereed

Fingerprint

Trajectories
community

Keywords

  • Emergence score (EScore)
  • Technological emergence
  • Text analytics
  • Topic modeling

Cite this

@article{24fbaf0a92d94c8e974634cd4e3e53ab,
title = "Evaluating technological emergence using text analytics: Two case technologies and three approaches",
abstract = "Scientometric methods have long been used to identify technological trajectories, but we have seldom seen reproducible methods that allow for the identification of a technological emergence in a set of documents. This study evaluates the use of three different reproducible approaches for identifying the emergence of technological novelties in scientific publications. The selected approaches are term counting technique, the emergence score (EScore) and Latent Dirichlet Allocation (LDA). We found that the methods provide somewhat distinct perspectives on technological. The term count based method identifies detailed emergence patterns. EScore is a complex bibliometric indicator that provides a holistic view of emergence by considering several parameters, namely term frequency, size, and origin of the research community. LDA traces emergence at the thematic level and provides insights on the linkages between emerging research topics. The results suggest that term counting produces results practical for operational purposes, while LDA offers insight at a strategic level.",
keywords = "Emergence score (EScore), Technological emergence, Text analytics, Topic modeling",
author = "Samira Ranaei and Arho Suominen and Alan Porter and Stephen Carley",
year = "2019",
month = "11",
day = "2",
doi = "10.1007/s11192-019-03275-w",
language = "English",
journal = "Scientometrics",
issn = "0138-9130",
publisher = "Springer",

}

Evaluating technological emergence using text analytics : Two case technologies and three approaches. / Ranaei, Samira; Suominen, Arho (Corresponding Author); Porter, Alan; Carley, Stephen.

In: Scientometrics, 02.11.2019.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Evaluating technological emergence using text analytics

T2 - Two case technologies and three approaches

AU - Ranaei, Samira

AU - Suominen, Arho

AU - Porter, Alan

AU - Carley, Stephen

PY - 2019/11/2

Y1 - 2019/11/2

N2 - Scientometric methods have long been used to identify technological trajectories, but we have seldom seen reproducible methods that allow for the identification of a technological emergence in a set of documents. This study evaluates the use of three different reproducible approaches for identifying the emergence of technological novelties in scientific publications. The selected approaches are term counting technique, the emergence score (EScore) and Latent Dirichlet Allocation (LDA). We found that the methods provide somewhat distinct perspectives on technological. The term count based method identifies detailed emergence patterns. EScore is a complex bibliometric indicator that provides a holistic view of emergence by considering several parameters, namely term frequency, size, and origin of the research community. LDA traces emergence at the thematic level and provides insights on the linkages between emerging research topics. The results suggest that term counting produces results practical for operational purposes, while LDA offers insight at a strategic level.

AB - Scientometric methods have long been used to identify technological trajectories, but we have seldom seen reproducible methods that allow for the identification of a technological emergence in a set of documents. This study evaluates the use of three different reproducible approaches for identifying the emergence of technological novelties in scientific publications. The selected approaches are term counting technique, the emergence score (EScore) and Latent Dirichlet Allocation (LDA). We found that the methods provide somewhat distinct perspectives on technological. The term count based method identifies detailed emergence patterns. EScore is a complex bibliometric indicator that provides a holistic view of emergence by considering several parameters, namely term frequency, size, and origin of the research community. LDA traces emergence at the thematic level and provides insights on the linkages between emerging research topics. The results suggest that term counting produces results practical for operational purposes, while LDA offers insight at a strategic level.

KW - Emergence score (EScore)

KW - Technological emergence

KW - Text analytics

KW - Topic modeling

UR - http://www.scopus.com/inward/record.url?scp=85074849404&partnerID=8YFLogxK

U2 - 10.1007/s11192-019-03275-w

DO - 10.1007/s11192-019-03275-w

M3 - Article

AN - SCOPUS:85074849404

JO - Scientometrics

JF - Scientometrics

SN - 0138-9130

ER -