Questionnaires contain often also questions, to which answers are given as free-form text. This is an important source of information in qualitative analyses, which has many other applications as well. Currently there are no well accepted methods to include natural language text into automated analysis of documents, but recent results with the use of the SOM give some promise to this direction as well.
If similar answers or documents can be selected into separate groups having a meaningful description, the answers could be analyzed as separate categories and those results could be added to the training set of the concluding SOM. An illustration of the complete model can be found in figure 8.
Figure 8: Complete analysis model.