Yesterday our summer project started at
Creating a new special purpose search service – basically from the rough idea to a working prototype – in 5 weeks seems a bit crazy but I am confident that we get there ;-) In certain areas we already have an idea how much pages we have to crawl and how much content we have to analyze.
It is interesting that it already now becomes apparent that user interface issues and system architecture decisions are closely linked. E.g. doing a meta search while the user is waiting requires some other content that we can present while the user is expecting the results.