Summary

Abstract

Summarizing the contents of a book is a matter of personal preferences. One way to obtain more objectivity is to use formal criteria for the identification of the most important findings. In Chap. 8, we introduced text mining, in particular, analysis methods based on the term document matrix for the detection of structure in text data. Hence, we thought that it is self-evident to use the text mining approach as a starting point for a summary. Specifically, we used the following procedure in order to acquire an overview on the contents in the different chapters:

• Definition of a corpus containing the eight chapters of the book.

• Cleaning the documents in the standard way by removal of stop words, punctuation, and numbers.

• Definition of two document term matrices: one with words and the other one with words and bigrams. For these two matrices, some stemming was done, mainly to clean plurals. Furthermore, some additional stop words were removed, mainly words in context of the examples in Chap. 8.

• For both matrices, term frequency-inverse document frequencies (TF-IDF) were calculated.

• Definition of comparison clouds each based on 60 terms.

• Calculation of topic maps of order 2–8 for the term document matrices with stem terms.