- From: Todd A. Mancini <todd.mancini@daxat.com>
- Date: Tue, 18 Feb 2003 17:34:07 -0500
- To: "'Michael Rys'" <mrys@microsoft.com>, <public-qt-comments@w3.org>
- Message-ID: <000d01c2d79d$d9842a70$0201a8c0@qodfathr>
Michael, No, that's what I thought it meant - that the system MUST supply support for stopword elimination. But why? I fail to see a good reason to mandate stopword support, but I understand that certain implementations MAY want/need to have stopword support (because they need it to perform). I don't believe the majority of users want stopwords (although many believe they need them, because they have been using engines which require them). My personal experiences indicate that having stopwords is not regarded as nice functionality, just necessary functionality for the 'older technology' full-text engines. I'd be more inclined to agree that stopwords should be a MUST if you could give me an example of a query which is more powerful to the user due to the existence of stopword support (but not ignoring other technologies, such as stemming, spelling, thesaurus, phrase detection, etc.). The use-cases for stopwords, in my opinion, highlight why stopwords are bad and why a user would want to use an engine without them. To be clear, my point is this: if I can create a fantastic full-text search engine that can handle queries which include common words such as 'the', it seems a little silly for me to then do extra development to add stopword support simply so I can be compliant with the specification. This extra development does not make my engine any more powerful, and I argue provides no benefit to my users. Just my $0.02. Regards, -Todd
Received on Tuesday, 18 February 2003 20:53:11 UTC