On 25/08/07, Oleg Bartunov <oleg@xxxxxxxxxx> wrote: > On Fri, 24 Aug 2007, Phoenix Kiula wrote: > > > Hi, > > > > We have big blobs of text (average 10,000 characters) in a database, > > from which we would like to discover the most often repeated words or > > phrases. Can tsearch be used for this kind of pattern search? I > > suppose it's Text Mining 101 sort of stuff, nothing complex. > > there is stat() function, see > http://www.sai.msu.su/~megera/wiki/Tsearch_V2_Notes > for more details. > It's not fast, so better to save results in a table Thanks. This seems to give words only. How about phrases? If words are so slow, I shudder to think how long phrase analysis would take -- it that is possible at all? ---------------------------(end of broadcast)--------------------------- TIP 4: Have you searched our list archives? http://archives.postgresql.org/
