One things divides knowledge and data: an aim. This also applies to customer data. Maybe this sounds obvious, but large amounts of data lead to the temptation of trying to mine it. Data mining is fashionable at the moment.
I remember TV-data analysis. Audiences allowed for their channel flicking to be recorded and their trash TV preferences to be analysed. Various mathematical methods were applied to the data. There was hardly any difference between viewers of different programmes, but one group stuck out. They were the ones who were interested in the news.
The search went deeper; content was analysed and the content of broadcasts was checked for intellectual exceptions. Many interesting hypotheses were established and scenarios compared. Then the intern was also allowed to say something. The intern had discovered that these broadcasts were on just before the news. Long faces everywhere. We found a group out the audience who did not switch off their TV when a programme had finished.
The group just fizzled away. We had stepped on our own feet when it came to data mining. Unfortunately too, it was absolutely free from pain.