date: February 3, 2016
Lutz Bornmann recently promoted a relatively new piece of software through the SIGMETRICS mailing list 1) that can be used to analyze cited references exported from Web of Science (WoS). The software, CRExplorer, is freely available 2). Here I'll follow along the example at that site to locate significant and important knowledge management literature and see if any specific texts stand out as major contributors to the literature. Specifically, what we'll do is search WoS, export the results (articles, etc.) and the references cited/listed in those results, and then analyze them (the cited references).
As with many such examples (and studies), the validity is partially based on the size and quality of the database (the data) and the search parameters used to retrieve the data (data collection). Like the example at the above site, I'll search the WoS Core Collection using
knowledge management (in quotes) as my query and limit the field of search to titles. I'll also only download the results and the cited references (the references listed in the articles that are returned) for the 500 most cited publications – mainly to limit the publication set and keep the data relatively sane (per the advice found in the CRExplorer PDF guide 3).
After importing the file downloaded from WoS into CRExplorer (CRE), we learn the following:
Within CRE, we'll further reduce the publication set by removing all references cited only once (singletons) and all references published before 1958, when Polanyi's Personal Knowledge was published. There are quite a few singletons, and after removing these, only a handful of titles published before 1958 are left. Still, they leave a rather long tail in the graph that's generated. With the singletons and pre-1958 publications removed, the graph looks like this:
The documentation at the CRE site is nice. I can use the CRE graph to highlight and zoom in on parts of the graph, and the corresponding cited references are highlighted in the table view in the right panel of the CR software. For instance, zooming in on the peak in the early 1990s reveals a few highly cited references, including:
Huber, G.P. (1991). Organizational learning: The contributing processes and the literatures. Organizational Science, 2, 88-115. doi:http://dx.doi.org10.1287/orsc.2.1.88
Among others. In 1995, the leading cited reference is to Nonaka and Takeuchi's book:
Nonaka, I., & Takeuchi, H. (1995). The knowledge-creating company. New York: Oxford University Press.
The highest peak is in 1998 and the highest cited reference is to another book:
Davenport, T. H. & Prusak, L. (1998). Working knowledge: How organizations manage what they know. Boston: Harvard Business School Press.
It's a nice tool that can be used not just for bibliometric research but also with identifying works for literature reviews, with developing syllabi, and such. Another benefit is that it does a nice job structuring the data that's downloaded from WoS, and this data can be exported as a CSV file, which is easier to deal with than WoS's default bibliographic format 4).