atic data array, PHP file plus opcode cache is the fastest way I could
find to get it. I imagine it holds for HHVM too, maybe even more so. So
if we're porting it to PHP it'd make sense to make LM files PHP arrays.
--
Stas Malyshev
smalys...@wikimedia.org
welcome of course.
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery
e's exact match to what I typed it should win.
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery
() in SearchEngine is recommended instead. For now,
these hooks will be supported by base SearchEngine implementation, but
not by CirrusSearch.
The task for this is https://phabricator.wikimedia.org/T121430, it also
links the patches as they are now (still work in progress).
--
Stas Malyshev
smalys
owned upon - which
usually means converting to epic or something like that, for epic pretty
it's ok to have a lot of points derived from subtasks.
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://l
enwiki for 0 brings up a
> disambig page with lots of reasonable candidate results.
That may be a bug. I wonder if we don't have if ($search) somewhere that
leads to it - since "0" is falsy in PHP, it may make it look like "". $a
!= "" would
note inside the task that it is
waiting for being deployed, so nobody resolves it. Then when it's
deployed it can be actually resolved.
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia
Hi!
> We could add an extra attribute to the graph, e.g.
> (number of minutes), to let graph extension update cache expiry.
I think that'd be a nice addition.
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
dis
t from Firefox 40 that they are essentially different
browsers, but if not, it'd be nice to have this grouping maybe.
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/li
Hi!
I plan to send a note to wikitech about SearchEngine prefix completion
refactoring we did for suggester recently. Here's how it would look
like: https://www.mediawiki.org/wiki/User:Smalyshev_(WMF)/Suggester
Please review and suggest fixes/additions.
--
Stas Malyshev
smalys...@wikimedia.org
se add your window there. As I see you want a 7 hour window,
>> overlapping with other windows shouldn't be a problem. But, if you ever
>> plan to do work that might impact more than just your own systems,
>> please do coordinate with me.
>>
>> Thanks,
>>
>> Greg
>&
Hi!
> As my final hurrah, I've released the data access guidelines used by
> the Discovery team in research and analysis on to Meta. It can be
> found at https://meta.wikimedia.org/wiki/Discovery/Data_access_guidelines
Thank you very much for getting this done!
--
Stas Malysh
Hi!
> Now impatiently waiting for AROUND implementation in wdq2sparql :-)
That should work now too :)
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listi
ors are fine.
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery
Hi!
> I would've called it search summary, but some people seem to use that to
> refer to the snippets under the title/link for a result.
DuckDuckGo calls it "instant answer". Not sure it's the best term out
there, but that's what they call it.
--
Stas Malyshev
smalys.
Hi!
> Another option for personal categorization: 'flags' - it's the
> phabricator equivalent of bookmarks. Flags are personal, not shared
> publicly.
>
> A bit more info about flags is at https://phabricator.wikimedia.org/T102812
Thanks, I didn't even know it existed!
--
of
packages and see how it goes.
I must note though that "Edit paths" interface is very hard to use -
having one scroll box without search to locate a repo. I filed T140713
for it.
Thanks,
--
Stas Malyshev
smalys...@wikimedia.org
__
vel
>> algorithms agains the currently used ones. Are there any query logs
>> existing which I can use for this purpose?
We do have query logs, but they are not publicly accessible for privacy
reasons. You may want to check this out though:
https://meta.wikimedia.org/wiki/Research:Unde
question of "where one can get logs of the search queries
for public use" AFAIK is "there's no way to get it without signing
papers and going through procedures".
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery
, please feel
welcome to share.
[1] https://www.mediawiki.org/wiki/Help:CirrusSearch#filetype
P.S. at least filetype:video and filetype:audio are intuitive :)
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@list
Hi!
> So, what do we call this thing? Horrible first attempts:
>
> * ltr-pipeline
> * learn-to-rank-pipeline
> * bob
> * cirrussearch-ltr
> * ???
rank forge?
ML ranking?
sorting hat? :)
LTR is kind of confusing - for me it's the opposite of RTL which is
handling Hebrew
, my creativity is temporarily exhausted, but I may be back :)
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery
ant to do it for CREDIT.
--
Stas Malyshev
smalys...@wikimedia.org
___
Discovery mailing list
Discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery
though :)
--
Stas Malyshev
smalys...@wikimedia.org
___
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery
gories-rdf.dblist - we may
consider adding more wikis to it. E.g. see:
https://phabricator.wikimedia.org/T194139
So which wikis need to be added?
--
Stas Malyshev
smalys...@wikimedia.org
___
Discovery mailing list
Discovery@lists.wikimedia.org
actually want to do :)
--
Stas Malyshev
smalys...@wikimedia.org
___
Discovery mailing list
Discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery
ow
> from OCLC, so if that's not possible, then, yeah, there's no point.
Ah yes, you can combine! Just call Mediawiki API from inside SPRARQL
query and combine with other clauses:
https://www.mediawiki.org/wiki/Wikidata_Query_Service/User_Manual/MWAPI
--
Stas Malyshev
sm
entially) I
decided not to pursue this for now. We'd have basically to duplicate the
work we've done in Mediawiki to compose proper Elastic queries, parse
results, etc. and the best we'd have is the same thing we already have
with Mediawiki API search. So I decided not to duplicate efforts f
esult, i.e.
either open source code (with reusable license, i.e. no patents banning
reuse etc.) or open publication with freely accessible algorithms and
outcomes (or both?) I don't think it would make sense for us to
cooperate if we'd be unable to benefit from the results.
--
Stas Malyshev
smaly
29 matches
Mail list logo