If you could analyse patterns of searches, it might lead you to identify the emergent knowledge topics, the knowledge gaps, and the "hot potatoes" in your organisation, your community, or in society.
We tried just this approach of analysing questions within a big community of practice recently. The queries to the community forum were already characterised into topics, because when you ask a question using the forum software, you have to choose which topic it is related to. So that saved us having to assign categories.
We divided these topics into four quadrants;
1. Topics where there were few questions, but each one got lots of answers. These tended to be areas of common knowledge, where most people knew the answer and only a few new people did not. For these topics, we could write guidelines, knowledge assets or faqs for the benefit of the new staff
2. Lots of questions, lots of answers. These were the important and evolving Knowledge topics where it was worth while setting up community meetings such as Knowledge Exchanges so that we could start to exchange and document best practice.
3. Lots of questions, few answers. These were the problem areas, where some more research or action learning was needed to start to develop solutions.
4. Few questions, few answers. Our assumption was that these are not particularly important areas, but that it was worth watching them in case they developed into problem areas.
This was a very useful analysis and led to a greater understanding of the important evolving and problem topics within the community.