CO-WORD-BASED SCIENCE MAPS OF CHEMICAL-ENGINEERING .1. REPRESENTATIONS BY DIRECT MULTIDIMENSIONAL-SCALING

In this paper we present the results of the first part of our study on science mapping. We discuss a new approach to co-word analysis in order to solve or to circumvent some of the problems mentioned in the earlier literature. The mapping technique is based on a ''direct'' application to the word co-occurrence data matrix of multidimensional scaling. The subject area of our co-word mapping approach is chemical engineering. We have chosen this field of scientific research because we found it tempting to bibliometrically map such a large, rather multidisciplinary and economically important field. Furthermore, it is important to assess bibliometric methods and techniques not only for basic research fields, but also for the more applied and engineering fields. We discuss maps based on three different datasets (top-journals, top-scientists, and a conference). Furthermore, we discuss the differences between maps based on title- and on abstract-words. Moreover, we experiment with different relational indices to obtain optimal contrasts between interesting and not-interesting features. We explore the possibilities to visualize changes over time in one display. An important part of the work is the evaluation of the maps by an extensive expert judgement.