Figure 3

In order to display graphically the similarity between sections regarding keyword content, we took the inverse of the average number of shared keywords (Table 2) as a measure of dissimilarity between sections, and we plotted it as a dendogram (using [19]).