John, You might take a look at a patch I submitted to lucene-dev this morning (http://issues.apache.org/bugzilla/show_bug.cgi?id=30429) . It's aimed at pseudo-relevance feedback, so it automatically adds terms from an initial answer set to a query, but of course you can do whatever you want with the terms after they've been ranked. To rank terms and find the best descriptors the Robertson Selection Value is employed.
Best regards, Rene > On Aug 1, 2004, at 10:25 PM, John Adam wrote: > > Is there a way to get most significant words of a document if i give a > document number. > -- NEU: WLAN-Router f�r 0,- EUR* - auch f�r DSL-Wechsler! GMX DSL = superg�nstig & kabellos http://www.gmx.net/de/go/dsl --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
