For some reason ezSearch doesn't pick up words with a dash in it. It's a swedish installation and for some reason ezSearch doesn't find the swedish word a-kassa. It finds kassa, but not a-.
I also tried to understand however ezSearch has some fuzzines built in to it. Does anyone know?
I think it may be something todo with the standard analyser. ezSearch uses examine which uses lucene.net this has concept of analysers they analyse content and depending on which analyser is used it does different things, with standard analyser all english stop words are removed everything is lower cased and i suspect - is being removed. When you query your search term is run through same analyser and goes through same process. So as a test I would first of all using the search tool built into v7 (if you are using v7) find document with that term and have a look at how its stored. There is a swedish analyser but that is written in java for lucene v3 see http://swestemmer.sourceforge.net/
Words with a dash character
For some reason ezSearch doesn't pick up words with a dash in it. It's a swedish installation and for some reason ezSearch doesn't find the swedish word a-kassa. It finds kassa, but not a-.
I also tried to understand however ezSearch has some fuzzines built in to it. Does anyone know?
Jerker,
I think it may be something todo with the standard analyser. ezSearch uses examine which uses lucene.net this has concept of analysers they analyse content and depending on which analyser is used it does different things, with standard analyser all english stop words are removed everything is lower cased and i suspect - is being removed. When you query your search term is run through same analyser and goes through same process. So as a test I would first of all using the search tool built into v7 (if you are using v7) find document with that term and have a look at how its stored. There is a swedish analyser but that is written in java for lucene v3 see http://swestemmer.sourceforge.net/
Regards
Ismail
is working on a reply...