On Fri, 2006-09-15 at 15:31 +0200, Mark Müller wrote: > I guess terms will only be took into the corpus when the search found > results at least once for that term (and removed if no more results were > found). > > Persisting the corpus has to be done, but should be no problem.
I use ObjectIn&OutputStream. > It should be possible to weight the terms by number of searches or number of > results too i guess. The code refered to does that. The thing is that is handles phrases and not just terms. Just as Google suggest. But if you only have a couple of queries a day this is code might not be for you. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
