It looks like a bug...
As Trey said it happens sometimes.
But for peripheral neuritis it's not the case, the highlighter decided
to highlight a partial match ( *peripheral* neuropathy ) instead of the
full match to the redirect.
I think that when there's a single match to a word in the title any
other matches (even better ones) to redirects, categories or headings
will be ignored.
We will have to review the highlighter configuration to fix this issue.
Le 30/11/2015 15:50, Trey Jones a écrit :
Though it isn't exactly the same, this reminds me
of our discussion
back in Aug/Sep about Corn and Maize, much of which is on Phabricator:
https://phabricator.wikimedia.org/T110571 .
The behavior you want happens sometimes. Searching for "usually" gives
as the first result: Convention (norm) (redirect from Usually)
"difficult": Difficulty (redirect from Difficult)
"corn", among its many odd behaviors, gives: Maize (redirect from
Maize corn).
I think there's some sort of ranking going on for redirects that can
put a partial match above an exact match (hence "Maize corn" instead
of "corn" as the redirect shown).
More relevant to your example, I think that if there's a match on one
or more search terms in the redirected-to title, it blocks
redirected-from titles from being shown. Searching for "neropathic"
gets a spelling correction to "neuropathic", which gives as the first
result: Peripheral neuropathy (redirect from Neuropathic).
David may know more about either of these cases off the top of his
head. (But maybe not, since it was a while ago, and I recall that it
was very messy, and we declined to work on it then.)
So, at the very least, the control group would at least sometimes have
a chance of getting the test behavior because we do that now under
certain circumstances.
Another option is to consider this a bug fix so that we have
consistent behavior whenever there's an exact match of the query to a
redirect title, regardless of partial matches to the redirected-to
title or higher-ranked redirects.
It would also be interesting to know how often queries get redirected
results, and how often they show the redirects—either in general, or
as part of any A/B test.
—Trey
Trey Jones
Software Engineer, Discovery
Wikimedia Foundation
On Sat, Nov 28, 2015 at 3:40 AM, Oliver Keyes <okeyes(a)wikimedia.org
<mailto:okeyes@wikimedia.org>> wrote:
So I was looking up information on peripheral neuritis[0] and I
accidentally mistyped it as "peripheral neuriti". The good news: the
autocorrector worked out I'd done it wrong, corrected it, and sent me
automatically to the right results. Yay![1]
But looking at the results I see a really obvious improvement we could
make that would definitely improve the user experience in this
scenario. See, if you look at the first article on the list you'll see
it's "Peripheral neuropathy". Why? Because peripheral neuritis
redirects to that. But the article header appears in the search
results as "Peripheral neuropathy", since that's the real title.
But it's not what I searched for. What I searched for was neuritis. Is
neuritis the same as neuropathy? I dunno, I'm a random reader. Is this
a good search result to click on? No idea.
What I'd love for us to do is run an A/B test with two conditions:
1. Users who search for a term which redirects to an article get the
current experience (control)
2. Users who search for a term which redirects to an article get the
article title in the search results claiming to be the redirect title
(test)
I bet this would really improve the clickthrough rate for this class
of searches. It would definitely improve the UX.
[0] I'm researching thalidomide. Long story.
[1]
https://en.wikipedia.org/w/index.php?search=peripheral+neuriti&title=Sp…
--
Oliver Keyes
Count Logula
Wikimedia Foundation
_______________________________________________
discovery mailing list
discovery(a)lists.wikimedia.org <mailto:discovery@lists.wikimedia.org>
https://lists.wikimedia.org/mailman/listinfo/discovery
_______________________________________________
discovery mailing list
discovery(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery