ERIC Number: ED167076
Record Type: Non-Journal
Publication Date: 1977-Oct
Reference Count: N/A
The Application of Clustering Techniques to Citation Data. Research Reports Series B No. 6.
Arms, William Y.; Arms, Caroline
This report describes research carried out as part of the Design of Information Systems in the Social Sciences (DISISS) project. Cluster analysis techniques were applied to a machine readable file of bibliographic data in the form of cited journal titles in order to identify groupings which could be used to structure bibliographic files. Practical applications of the data obtained from clusters of journal title citations were to include the planning patterns of journal coverage for secondary services, ascertaining information on the structure of the primary literature of a discipline, and the provision of data to help validate classification schemes and thesauri. The analyses were carried out on a file of over 60,000 citations taken from social science journals, and a special cluster analysis was made of citations taken from criminology source journals at three different time intervals. Included here are a review of clustering techniques applicable to bibliographic data, a detailed discussion of the results, and an evaluation of the practical value of the technique. It was concluded that the use of cluster analysis is probably not a very practical aid to the design of secondary services in the social sciences, though the techniques nevertheless have important bibliographic applications elsewhere. A bibliography is appended. (Author/JD)
Publication Type: Reports - Research; Reports - Evaluative
Education Level: N/A
Authoring Institution: Bath Univ. of Technology (England). Univ. Library.