As no statistical time series was available for the new refined, “split” concepts, the next goal of the research was generating a time series for all 33 factors, based on information in the literature, relying on the available expert judgment. Relying only on expert knowledge would have led to rather unreliable results, thus, an objective method for analysing the texts of the literature had to be proposed. Text mining seemed to be the easiest way to go, especially, as it promised to be the most suitable way to extract the time series in the time period covered by these literatures for all the new concepts. We assumed that this approach would allow the extraction of structured knowledge from the available unstructured or only slightly structured text files. Thus, the next step was to gather and transform the textual knowledge and information from these different document sources with applying proper machine intelligence. A specialised text mining expert company offered help as “sub-contractors” to the team’s research.