Measuring the diffusion of an innovation: A citation analysis

Abstract

Innovations transform our research traditions and become the driving force to advance individual, group, and social creativity. Meanwhile, interdisciplinary research is increasingly being promoted as a route to advance the complex challenges we face as a society. In this paper, we use Latent Dirichlet Allocation (LDA) citation as a proxy context for the diffusion of an innovation. With an analysis of topic evolution, we divide the diffusion process into five stages: testing and evaluation, implementation, improvement, extending, and fading. Through a correlation analysis of topic and subject, we show the application of LDA in different subjects. We also reveal the cross-boundary diffusion between different subjects based on the analysis of the interdisciplinary studies. The results show that as LDA is transferred into different areas, the adoption of each subject is relatively adjacent to those with similar research interests. Our findings further support researchers' understanding of the impact formation of innovation.

The impact of Chinese library and information science on outside disciplines: A citation analysis

“…First, citations still play a major role in the reliable measurement of research quality (Guerrero-Bote et al, 2007) and its impact (Bornmann and Daniel, 2008). Since publications are admitted to the academic community via peer review, citations, especially those in peer-reviewed journals, are associated with high quality and credibility (Zhai et al, 2018). Citations are at least a partial acknowledgement of the intellectual influence of cited documents on the citing documents (Cole and Cole, 1972; Thornley et al, 2015).…”

“…Second, if tacit knowledge flows play a role in complementing explicit knowledge flows rather than substituting for them, citations may capture at least part of such flows (Orazbayev, 2017). Third, citations are ‘frozen footprints in the landscape of scholarly achievement’ (Cronin, 1981), document the diffusion process (Zhai et al, 2018) and imply the knowledge flows across various disciplines (Yan, 2014). Citations, a common indicator of knowledge flows, are evaluated empirically by a large number of quantitative studies in interdisciplinary knowledge transfer (Hassan and Haddawy, 2013; Rodríguez, 2017; Sugimoto et al, 2008; Yan, 2014).…”

“…Citations, a common indicator of knowledge flows, are evaluated empirically by a large number of quantitative studies in interdisciplinary knowledge transfer (Hassan and Haddawy, 2013; Rodríguez, 2017; Sugimoto et al, 2008; Yan, 2014). Finally, with the increasing ease of access to large publication databases that capture major scientific activities (Zeng et al, 2017) and cumulative citation data over a long period (Zhai et al, 2018), new opportunities are created for multi-level/multi-dimensional analysis. This data-driven approach tracks and quantifies the permeable process of knowledge export (Yan et al, 2013) using rich mathematical and computational models (Zeng et al, 2017).…”

“…In scientific communities, knowledge import shows the influence of knowledge from outside disciplines (Chen et al, 2018), while knowledge export is an indicator of the scientific strength and impact needed to cross the traditional boundaries of the home discipline (Wormell, 1998). The effective adoption and diffusion of knowledge are the keys to maintaining the value of research achievements and bringing that knowledge into full play (Zhai et al, 2018). The international library and information science (LIS) community has made a strong effort to demonstrate knowledge export from LIS and to explore its external intellectual impact (Cronin and Meho, 2008; Larivière et al, 2012).…”

“…The strengths of the method, however, lie in the following aspects. First, citations included in peer-reviewed publications can be regarded as a credible proxy (Zhai et al, 2018) that partially represents the intellectual impact of knowledge export (Thornley et al, 2015). Second, in scholarly communications, knowledge is disseminated primarily by reading, adopting and citing scientific publications (Zeng et al, 2017); therefore, citations may capture aspects of the main scientific knowledge flows (Orazbayev, 2017).…”