[09:00:11] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Chtnnh) @Sumit Do you have any suggestions for the tests I could run on this? I am seem to be lost as to how to proceed [10:26:19] 10MediaWiki-extensions-ORES, 10Scoring-platform-team, 10Discovery-Search, 10NewcomerTasks 1.1, and 3 others: Expose ORES drafttopic data in ElasticSearch via a custom CirrusSearch keyword - https://phabricator.wikimedia.org/T240559 (10kostajh) For tech news: > a new search keyword, `articletopic:` is ava... [10:54:48] 10Scoring-platform-team, 10Discovery-Search, 10NewcomerTasks 1.1, 10Growth-Team (Current Sprint): Once the ORES articletopic - ElasticSearch pipeline is set up, update data about all articles - https://phabricator.wikimedia.org/T243357 (10kostajh) [10:57:28] 10Scoring-platform-team, 10Discovery-Search, 10NewcomerTasks 1.1, 10Growth-Team (Current Sprint): Once the ORES articletopic - ElasticSearch pipeline is set up, update data about all articles - https://phabricator.wikimedia.org/T243357 (10kostajh) >>! In T243357#5924365, @EBernhardson wrote: > A manual all... [13:53:21] o/ [14:21:49] 10Scoring-platform-team, 10Wikilabels, 10articlequality-modeling, 10artificial-intelligence: Build article quality model for ptwikipedia - https://phabricator.wikimedia.org/T246663 (10GoEThe) [14:38:02] 10Scoring-platform-team, 10Wikilabels, 10editquality-modeling, 10artificial-intelligence: Build draft quality model for ptwikipedia - https://phabricator.wikimedia.org/T246667 (10GoEThe) [14:42:14] 10Scoring-platform-team, 10Wikilabels, 10editquality-modeling, 10artificial-intelligence: Edit quality campaign for ptwikipedia - https://phabricator.wikimedia.org/T246668 (10GoEThe) [15:48:49] 10Scoring-platform-team, 10Wikilabels, 10editquality-modeling, 10artificial-intelligence: Edit quality campaign for ptwikipedia - https://phabricator.wikimedia.org/T246668 (10Halfak) Hi @GoEThe! We already have some edit quality models for ptwiki. See https://ores.wikimedia.org/v3/scores/ptwiki See also... [16:05:14] 10ORES, 10Scoring-platform-team, 10patch-welcome: Extend the GUI for ORES - https://phabricator.wikimedia.org/T107723 (10Halfak) @Chtnnh, I think the right next step here is to get some proposed changes together so that we can prioritize. You can look at the current version of the UI here: https://ores.wiki... [16:16:47] 10Scoring-platform-team, 10Wikilabels, 10articlequality-modeling, 10artificial-intelligence: Build article quality model for ptwikipedia - https://phabricator.wikimedia.org/T246663 (10Halfak) Looks like you can find the {{[marca de projeto](https://pt.wikipedia.org/w/index.php?title=Especial:P%C3%A1ginas_a... [16:17:20] 10Scoring-platform-team, 10Discovery-Search, 10NewcomerTasks 1.1, 10Growth-Team (Current Sprint): Once the ORES articletopic - ElasticSearch pipeline is set up, update data about all articles - https://phabricator.wikimedia.org/T243357 (10EBernhardson) I've also written everything necesssary to do the same... [16:18:15] 10Scoring-platform-team, 10Wikilabels, 10articlequality-modeling, 10artificial-intelligence: Build article quality model for ptwikipedia - https://phabricator.wikimedia.org/T246663 (10Halfak) Oh! I see that the first number is an *automatic evaluation*. [16:26:03] 10Scoring-platform-team, 10Wikilabels, 10editquality-modeling, 10artificial-intelligence: Build draft quality model for ptwikipedia - https://phabricator.wikimedia.org/T246667 (10Halfak) We're mostly looking to catch "spam", "vandalism", and other types of problematic drafts that are obvious from the conte... [17:09:24] 10ORES, 10Scoring-platform-team, 10patch-welcome: Extend the GUI for ORES - https://phabricator.wikimedia.org/T107723 (10Chtnnh) @Halfak Yeah I think you're right on that. Where can I find the code for the current GUI? [17:36:44] 10ORES, 10Scoring-platform-team, 10patch-welcome: Extend the GUI for ORES - https://phabricator.wikimedia.org/T107723 (10Chtnnh) Here's a list of ideas I could come up with: 1. Remove the predictions below a certain magnitude 2. Sort predictions in order of magnitude 3. Add visualization, graphs or pie char... [17:44:47] 10ORES, 10Scoring-platform-team (Current), 10editquality-modeling, 10artificial-intelligence: Visualize the relationship between the probability of reversion and ores scores - https://phabricator.wikimedia.org/T224918 (10Halfak) Moving this to review because I have been reviewing @Groceryheist's work. I t... [18:06:33] 10Jade, 10Scoring-platform-team (Current): Handle empty Jade pages - https://phabricator.wikimedia.org/T246033 (10ACraze) After some initial research, it seems that the TitleIsAlwaysKnown hook is not what we need. Maybe we can reverse engineer https://www.mediawiki.org/wiki/Extension:Special404 to get somethi... [18:07:00] o/ chtnnh [18:07:42] 10Jade, 10Scoring-platform-team (Current): Handle empty Jade pages - https://phabricator.wikimedia.org/T246033 (10ACraze) [18:10:39] halfak so I had a few question regarding testing the models [18:10:55] as I have not worked with codebases of this size previously [18:11:33] when I modified the parsed.py in the revscoring repo, @sumit suggested I run some local tests [18:11:41] but I was lost on how to proceeed [18:12:01] Right so. We use "pytest". It runs the tests in the "tests/" directory. [18:12:12] aha [18:12:27] is there a readme somewhere that I can refer to [18:12:42] You can find the tests related to parsed.py in https://github.com/wikimedia/revscoring/blob/master/tests/features/wikitext/tests/test_parsed.py [18:13:07] https://github.com/wikimedia/revscoring#running-tests [18:13:28] so I just find the relevant test file for all files I have modified and call those test files with python3 [18:14:47] You'll want to run them with "pytest" [18:15:03] e.g. "pytest tests/features/wikitext/tests/test_parsed.py" [18:15:12] that will just run that specific test file. [18:15:24] aha okay [18:15:27] If you run it with no argument, it will run all of the tests. [18:15:36] oh and that will take time [18:22:52] okay now that testing is sorted [18:22:57] about the tasks, [18:23:14] you think I should move on to tasks that are not listed as good first tasks about now? [18:30:28] Right. I've been thinking about suggesting one to you that I think could be interesting. [18:30:36] sure go ahead [18:30:41] https://phabricator.wikimedia.org/T246663 [18:31:00] btw, all tests passed. only 2 unrelated warning [18:32:38] halfak what is ptwikipedia? [18:34:52] Portuguese Wikipedia [18:34:53] task does seem interesting [18:35:05] classifier for article quality [18:35:13] relates with our work in the other task [18:35:21] but for a different language altogether [18:37:09] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Chtnnh) @Sumit Thanks for the suggestion! No errors thrown, PR has been submitted [18:39:28] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Halfak) Could you link to the PR? [18:49:23] 10Scoring-platform-team, 10articlequality-modeling, 10artificial-intelligence: Add text complexity scoring to article quality models - https://phabricator.wikimedia.org/T246438 (10Chtnnh) https://github.com/wikimedia/revscoring/pull/474 [18:51:41] halfak I made changes to my PR [18:51:48] Here's a link: https://github.com/wikimedia/articlequality/pull/102 [18:58:00] Nice work haksoat. I just added some notes. [19:03:20] Okay halfak I also made comments now. Help check when chanced. [19:03:50] chtnnh: since you're adding a new datasource, you should also manually call that datasource to check if you're indeed getting sections extracted from the revision. If you do api_extractor.extract(revid, wikitext.revision.datasources.sections) it should give you sections [19:04:00] see https://github.com/wikimedia/revscoring/blob/59dda23e10dcda45ce5fe82af5035fa92ba3de01/examples/extraction.py [19:08:05] codezee let me test that out this time [19:08:10] thanks so much :) [19:08:37] \o/ [19:09:48] wikimedia/articlequality#276 (datasource_cleanup - 878bcee : halfak): The build passed. https://travis-ci.org/wikimedia/articlequality/builds/657433733 [19:37:08] I need to break for lunch. I'll be back in a bit. [20:38:58] 10ORES, 10Scoring-platform-team (Current), 10Operations: ORES uwsgi consumes a large amount of memory and CPU when shutting down (as part of a restart) - https://phabricator.wikimedia.org/T242705 (10Halfak) @akosiaris, I've run a couple of tests now using threads rather than process-based workers and the beh... [22:39:32] I'm heading out for the day. Take care folks!