On Tue, Jul 19, 2011 at 12:20 PM, <caraman.mi...@gmail.com> wrote:
> Israel, if you have this implemented, I'd appreciate if you can crunch some > numbers so I know how slow it actually is, for future comparison? Let's say > on 100.000 results, each of which have up to 50 words, or 50.000 results > with 100 words each ... how much time does it take or how many queries per > second can it handle. Do you have your implementation as open source > somewhere? > > It was very, very slow. You should probably run your own tests to verify, but as I remember we had to limit the amount of documents read to 5000 to keep response time to less than a minute. Unfortunately, I don't have any source code to offer.
