Mixed-indicators model for identifying emerging research areas

Abstract

This study presents a mixed model that combines different indicators to describe and predict key structural and dynamic features of emerging research areas. Three indicators are combined: sudden increases in the frequency of specific words; the number and speed by which new authors are attracted to an emerging research area, and changes in the interdisciplinarity of cited references. The mixed model is applied to four emerging research areas: RNAi, Nano, h-Index, and Impact Factor research using papers published in the Proceedings of the National Academy of Sciences of the United States of America (1982) and in Scientometrics (1978. Results are compared in terms of strengths and temporal dynamics. Results show that the indicators are indicative of emerging areas and they exhibit interesting temporal correlations: new authors enter the area first, then the interdisciplinarity of paper references increases, then word bursts occur. All workflows are reported in a manner that supports replication and extension by others.

Knowledge Networks of the Information Technology Management Domain: A Social Network Analysis Approach

“…Whitley & Galliers (2007) Important IS-related bibliometric studies (without the SNA component) include Mutschke & Haase (2001), who used socio-cognitive analysis to examine the relationship exists between actors' positions in scientific networks and the innovativeness of the themes they examine, Cunningham & Dillon (1997), who examined the patterns of authorship in five information systems journals, and Whitley & Galliers (2007), who identified the most frequently cited texts in IS literature (Whitley & Galliers, 2007). Other studies identified in the literature include Polites and Watson (2009), who examined the relationship among IS journals, and Guo et al (2011), who used a scientometric approach to identify emerging research areas.…”

An analysis of the information technology outsourcing domain: A social network and Triple helix approach

“…The burst detection algorithm “employs a probabilistic automaton whose states correspond to the frequencies of individual words and state transitions correspond to points in time around which the frequency of the world changes significantly. Given a set of time stamped text, e.g., abstracts and publication years of the papers, the algorithm identifies those abstract words that experience a sudden increase in usage frequency and outputs a list of these words together with the beginning and ending of the burst and the burst strength that indicates the change in usage frequency” (Guo et al., , pp. 422–423).…”

International trade negotiation analysis: network and semantic knowledge infrastructure

“…Prior to the burst detection routine being applied, all data were lowercased and tokenized while common stop words were also removed. Having been extensively used (Kleinberg 2003) to identify emerging trends in their research domains (Mane and Borner 2004;Chen 2006;Chen et al 2009;Guo et al 2011) the burst technique forms a key component of this research.…”