Publication - Exploiting keyword life cycle to analyse the dynamics | Sciences, technologies et visualisations

Exploiting keyword life cycle to analyse the dynamics

Research Area: Text analytics Year: 2010
Type of Publication: In Proceedings
Authors:
  • Aurélie Delemarle
  • Audrey Baneyx
  • Philippe Brucker
  • Philippe Laredo
  • Bernard Kahane
  • Lionel Villard
Book title: 3rd network of indicators designers conference Number: 3
Series: PRIME-ENID Pages: 3
Address: Paris, France
Organization: CNAM Month: March
Abstract:
We argue that it is possible to follow the dynamics of a scientific field as it emerges. Being able to do such allows for example, industrial actors to position themselves regarding their competences and public policy makers to support the development of the field. For that matter, we follow traces of its emergence through publications analysis (Bonaccorsi and Thoma, 2005; Bozeman et al., 2007; Mangematin et al, 2008). Data included in large databases and extracted from the WoS brings us the possibility to characterise a scientific field. The use of traditional methods such as journal or co-journals analysis show the main journals of a field and the relationships between them (Hirsch, 2005, van Raan, 2003). The dynamics of journals and the co-citation indicator can illustrate the structuration of a field (Leydesdorff and Rafols, 2009; Park and Leydesdorff, forthcoming, Leydesdorff and Schank, 2008, Leydesdorff, 2007 ). Co-authors analysis (Callon et al, 1986; White and McCain, 1998) points to the networks of relationships between authors. It has been used to illustrate the internal structure of a field and its central and most powerful actors. Last, co-word analysis used on title and/or abstract points to the science itself showing how elements of a knowledge bases are articulated (van den Besselaar and Heimeriks, 2006). However, we consider that all these traditional methods used to illustrate the organisation of a scientific field can only be used as it is already structured. Indeed, journal analysis is not possible for an emerging field as it does not yet have reference journals (see also Zitt and Bassecoulard, 2006). Co-authors analysis cannot be used either as a community of scientists with star scientists (Darby and Zucker, 2006) does not yet exist. Last, co-word analysis cannot be used as the field is not defined yet which makes it impossible to choose which words should be selected and which should not. Following Bonaccorsi (2007), we argue that a keywords analysis can illustrate the dynamics of an emerging field. Bonaccorsi showed that an emerging field has a high rate of new keyword introduction. The technological plateform CorTexT (IFRIS1) developed a tool named Kwords Lab to explore the keywords of an emerging field. On the contrary to past studies on top of keywords from the abstract or the title, we also take into consideration keywords provided by individual authors as well as journal keywords. If this choice enriches our analysis it also brings heterogeneity which need to be dealt with.