Page MenuHomePhabricator
Create Task
Maniphest T282583

Searching in non-English languages doesn't return expected results
Closed, ResolvedPublic
Actions

Assigned To
Cparle
Authored By
Cparle
May 11 2021, 4:00 PM
Tags
Referenced Files
F34488796: Screen Shot 2021-06-09 at 4.58.36 PM.png
Jun 10 2021, 12:08 AM
F34488793: Screen Shot 2021-06-09 at 4.36.06 PM.png
Jun 10 2021, 12:08 AM
F34488778: Screen Shot 2021-06-09 at 4.30.00 PM.png
Jun 9 2021, 11:41 PM
F34472448: Screen Shot 2021-05-27 at 12.10.47 PM.png
May 27 2021, 7:15 PM
F34472450: Screen Shot 2021-05-27 at 12.12.16 PM.png
May 27 2021, 7:15 PM
F34472442: Screen Shot 2021-05-27 at 12.04.53 PM.png
May 27 2021, 7:15 PM
F34472418: Screen Shot 2021-05-27 at 11.48.42 AM.png
May 27 2021, 7:15 PM
F34472415: Screen Shot 2021-05-27 at 11.48.04 AM.png
May 27 2021, 7:15 PM
View All 12 Files
Subscribers
Aklapper
Cparle
Etonkovidova
JeanFred
Keegan

Description

For example

https://commons.wikimedia.org/w/index.php?search=chien&title=Special:MediaSearch&go=Go&type=image&uselang=fr

contains a lot of images with 'chien' in the title before we get any images of dogs, despite the uselang=fr param. statement matches are rated more highly than title matches, so that's unexpected

Details

SubjectRepoBranchLines +/-
Remove 'wikibase' profile, add lang to cache keymediawiki/extensions/WikibaseMediaInfomaster+9 -5
Customize query in gerrit

Event Timeline

Cparle created this task.May 11 2021, 4:00 PM
Restricted Application added a subscriber: Aklapper. · View Herald TranscriptMay 11 2021, 4:00 PM
Cparle claimed this task.May 11 2021, 4:00 PM
Cparle moved this task from Incoming to Doing on the Structured-Data-Backlog (Current Work) board.
Keegan added a subscriber: Keegan.May 11 2021, 4:06 PM
JeanFred added a subscriber: JeanFred.May 11 2021, 7:28 PM
gerritbot added a comment.May 13 2021, 10:43 AM

Change 690403 had a related patch set uploaded (by Cparle; author: Cparle):

[mediawiki/extensions/WikibaseMediaInfo@master] Remove 'wikibase' profile, add lang to cache key

https://gerrit.wikimedia.org/r/690403

gerritbot added a project: Patch-For-Review.May 13 2021, 10:43 AM
Cparle added a comment.May 13 2021, 10:44 AM

Comparison of results from the analysis tool before and after this change

before change:

F1 Score      | 0.59321010027788
Precision@10  | 0.82402707275804
Precision@25  | 0.79368213228036
Precision@50  | 0.76342624065262
Precision@100 | 0.72581429265921
Recall        | 0.53000863557858

after change:

F1 Score      | 0.59456193353474
Precision@10  | 0.82993197278912
Precision@25  | 0.79860418743769
Precision@50  | 0.76543209876543
Precision@100 | 0.72519455252918
Recall        | 0.53108808290155

... so all scores are slightly improved

Cparle moved this task from Doing to Code Review on the Structured-Data-Backlog (Current Work) board.May 13 2021, 10:55 AM
matthiasmullie moved this task from Code Review to Needs QA on the Structured-Data-Backlog (Current Work) board.May 25 2021, 9:12 AM
gerritbot added a comment.May 25 2021, 9:27 AM

Change 690403 merged by jenkins-bot:

[mediawiki/extensions/WikibaseMediaInfo@master] Remove 'wikibase' profile, add lang to cache key

https://gerrit.wikimedia.org/r/690403

ReleaseTaggerBot added a project: MW-1.37-notes (1.37.0-wmf.9; 2021-06-07).May 25 2021, 10:00 AM
Maintenance_bot removed a project: Patch-For-Review.May 25 2021, 10:10 AM
Etonkovidova moved this task from Needs QA to Verify on Production on the Structured-Data-Backlog (Current Work) board.EditedMay 27 2021, 7:15 PM
Etonkovidova added a subscriber: Etonkovidova.

For comparative testing on commons wmf.9

commons wmf.7commons wmf.9
search term "chien"search term "chien"
French UI: https://commons.wikimedia.org/w/index.php?search=chien&title=Special:MediaSearch&go=Go&type=image&uselang=fr
Screen Shot 2021-05-27 at 11.48.04 AM.png (1×2 px, 3 MB)
The second image has dogs
Screen Shot 2021-06-09 at 4.30.00 PM.png (1×2 px, 3 MB)
English UI: https://commons.wikimedia.org/w/index.php?search=chien&title=Special:MediaSearch&go=Go&type=image
Screen Shot 2021-05-27 at 11.48.42 AM.png (1×2 px, 2 MB)
no changes (practically) in the ranking of the images
Screen Shot 2021-06-09 at 4.36.06 PM.png (1×2 px, 3 MB)
Russian UI: https://commons.wikimedia.org/w/index.php?search=chien&title=Special:MediaSearch&go=Go&type=image&uselang=ru
Screen Shot 2021-05-27 at 11.55.14 AM.png (1×2 px, 3 MB)
no changes (practically) in the ranking of the images
Screen Shot 2021-06-09 at 4.58.36 PM.png (1×2 px, 3 MB)
search term "chat"search term "chat"
French UI: https://commons.wikimedia.org/w/index.php?search=chat&title=Special:MediaSearch&go=Go&type=image&uselang=fr
Screen Shot 2021-05-27 at 12.02.42 PM.png (1×2 px, 3 MB)
no changes
English UI: https://commons.wikimedia.org/w/index.php?search=chat&title=Special:MediaSearch&go=Go&type=image&uselang=en
Screen Shot 2021-05-27 at 12.04.53 PM.png (1×2 px, 3 MB)
no changes
Russian UI: https://commons.wikimedia.org/w/index.php?search=chat&title=Special:MediaSearch&go=Go&type=image&uselang=ru
Screen Shot 2021-05-27 at 12.00.58 PM.png (1×2 px, 2 MB)
no changes
search term "кот" (Russian for "cat"search term "кот" (Russian for "cat")
French UI: https://commons.wikimedia.org/w/index.php?search=%D0%BA%D0%BE%D1%82&title=Special:MediaSearch&go=Go&type=image&uselang=fr
Screen Shot 2021-05-27 at 12.10.47 PM.png (1×2 px, 3 MB)
no changes
English UI: https://commons.wikimedia.org/w/index.php?search=%D0%BA%D0%BE%D1%82&title=Special:MediaSearch&go=Go&type=image&uselang=en
Screen Shot 2021-05-27 at 12.06.21 PM.png (1×2 px, 3 MB)
no changes
Russian UI: https://commons.wikimedia.org/w/index.php?search=%D0%BA%D0%BE%D1%82&title=Special:MediaSearch&go=Go&type=image&uselang=ru
Screen Shot 2021-05-27 at 12.12.16 PM.png (1×1 px, 2 MB)
no changes

The improvement is noticeable for the search term chien when French language is set for the UI language; all other cases were checked to see if there are no regression - all seems to be good.

Etonkovidova closed this task as Resolved.Jun 10 2021, 12:09 AM
Content licensed under Creative Commons Attribution-ShareAlike (CC BY-SA) 4.0 unless otherwise noted; code licensed under GNU General Public License (GPL) 2.0 or later and other open source licenses. By using this site, you agree to the Terms of Use, Privacy Policy, and Code of Conduct. · Wikimedia Foundation · Privacy Policy · Code of Conduct · Terms of Use · Disclaimer · CC-BY-SA · GPL