lucene-openrelevance-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mark Bennett <>
Subject Re: Search Engine Vendors (was: or-user perspective, teams, etc (was: Comments on ORP Wiki Additions ?))
Date Fri, 12 Feb 2010 21:47:14 GMT
Hi Robert,

I'm gonna need to get some more coffee...

* comment below

On Fri, Feb 12, 2010 at 11:47 AM, Robert Muir <> wrote:

> in general i don't see why it has to be a competition anyway, e.g. i don't
> waste time comparing any relevance scores from lucene to other search
> engines really, i do look at some published results to make sure i am in the
> right ballpark, but i guess how good or bad someone else does is N/A.

* Understood, I'm usually like that too.  But folks who are looking to spend
a half million dollars need their due diligence, and this is just one of the

>  If so, that's a perfectly logical and defensible stance!  Sadly it's not
>> a my company can adopt.  The reality is we need to interact with the top
>> players whether they are reasonable or not.
> I do not think this is necessarily the case as far as relevance goes, that
> commercial companies are the 'top players'.  Do you have any evidence to
> support this?
* Ah, my fault again.  The "we" in that context was my company, not ORP.

Wrt "evidence", not sure which point given my ambiguous use of "we", but
just in case:

Either in official PR, and/or as a feature sets discussed with potential
customers, virtually all vendors talk about how wonderful their relevancy
is, and I've seen that since the early 90's.  And when we see pitches for
new companies, it's mentioned by about half of them.  Thankfully other
startups are now talking about doing more work with less hardware and
maintenance, woo hoo!  There are plenty of other perennial themes in
marketing collateral and RFPs besides relevance of course.

> But the GOOD NEWS is that ORP might offer some real benefits to them, and
>> some vendors are really quite reasonable and logical at times.
> ...
> The main reason I don't want us to be biased towards lucene, is so that
> results are not biased towards lucene's status quo, else we won't want to
> able to properly judge potential improvements. So I really do care a lot
> about this.
* Here here!!!

Once we all get settled in, I'd also like to talk about search and relevancy
beyond the "Text-In-Text-Out" model.  The entire industry has been stuck in
TITO-land for far too long, but that's a subject for another day.

> if you know of a way to sucker in non-ridiculous commercial guys into
> contributing stuff to JIRA, then please proceed :)
* Ha!  I love it.

Some of them won't require any such subterfuge.  Though I certainly don't
speak for them, I'd certainly think Google and FAST/Microsoft to be at least
somewhat interested.  Harder to predict the reaction of the other big
vendor, who already *knows* they're the best.  ;-)  I suspect some of the
Tier-2 folks would be interested too.

View raw message