Mining text archives: creating readable maps to structure and describe document collections

摘要

With the ever-growing amount of unstrutured textual data on the web,mining these text collections is of increasing importance ofr the understanding of document archives.Particularly the self-organizing map has shown to be very well suited for this task.however,the interpretation of the resulting document maps still requires a tremendous effort,especially as far as the analysis of the features learned and the characteristics of identified text clusters are concerned learned and the characteristics of identified text clusters are concerned.In this paper we present the LabelSOM method which,based on the features learned by the map,automatically assigns a set of keywords to the units of the map to describe the concepts of the underlying text clusters,thus making the characteristics of the various topical areas on the map explicit.