[01:49:44] 10Scoring-platform-team (Current), 10Research Ideas, 10artificial-intelligence: [Epic] Paid editing (COI) detection model - https://phabricator.wikimedia.org/T120170#4266414 (10Doc_James) All the details being feed in are undisclosed paid promotional editors. Thus it will be geared towards the undisclosed ty... [14:04:47] o/ [14:05:04] ewhit_, thanks for getting back to me with your PDP material :) [14:05:21] Thanks for taking a look at it! [14:14:13] 10Scoring-platform-team (Current), 10Research Ideas, 10artificial-intelligence: [Epic] Paid editing (COI) detection model - https://phabricator.wikimedia.org/T120170#4267111 (10Halfak) Re. "disclosed vs. undisclosed" -- I think disclosure is a problem that is better handled by structured data than modeling.... [15:09:27] 10Scoring-platform-team (Current), 10Wikilabels, 10articlequality-modeling, 10artificial-intelligence: Train/test article quality model for euwiki - https://phabricator.wikimedia.org/T171119#4267260 (10Halfak) Sorry for the delay. I'm snowed under with end-of-fiscal year stuff at the WMF. We're looking t... [19:29:23] Solid essay ewhit_. You should consider posting this publicly :D [19:29:55] thank you! I can post it on medium after I tweak the beginning a bit more (too similar to my op-ed, which I agreed to only publish in one place) [19:33:19] 10Scoring-platform-team, 10editquality-modeling, 10User-Zoranzoki21, 10artificial-intelligence: Train / test reverted model for srwiki - https://phabricator.wikimedia.org/T194745#4267975 (10Acamicamacaraca) [19:33:21] 10Scoring-platform-team (Current), 10Bad-Words-Detection-System, 10revscoring, 10User-Ladsgroup, 10artificial-intelligence: Add language support for Serbian - https://phabricator.wikimedia.org/T174687#4267974 (10Acamicamacaraca) 05Resolved>03Open [19:33:55] halfak: I'm looking into options for transcribing the interviews. I think probably the cheapest thing to do is have me do it. I will need to buy a ~$25 transcription software license (I have an account at UM that can cover this). All options to outsource transcription are more expensive, at around ~$1-2 per minute, and my interviews right now have been 30-60 min long. Should I go ahead and get the transcription software [19:33:56] license? [19:35:06] 10Scoring-platform-team (Current), 10Bad-Words-Detection-System, 10revscoring, 10User-Ladsgroup, 10artificial-intelligence: Add language support for Serbian - https://phabricator.wikimedia.org/T174687#3570233 (10Acamicamacaraca) I reopened task because all edits in campaign are reviewed and this need to... [19:35:09] brb meeting [19:38:14] 10Scoring-platform-team (Current), 10Bad-Words-Detection-System, 10revscoring, 10User-Ladsgroup, 10artificial-intelligence: Add language support for Serbian - https://phabricator.wikimedia.org/T174687#4267981 (10Acamicamacaraca) @Ladsgroup can you enable this? Thanks in advance! [20:01:56] Back! [20:02:00] * halfak reads backscroll [20:02:38] ewhit_, if you want to use a service, get me an estimate of the cost and I'll find out if we can swing it. [20:03:10] If you'd rather just do it, then go ahead and get the software. I'll approve the cost. [20:06:59] halfak|Network_d: where's the best place to look to understand the ru.wiki ORES wp10 scale? [20:07:40] ragesoss, https://ru.wikipedia.org/wiki/%D0%9F%D1%80%D0%BE%D0%B5%D0%BA%D1%82:%D0%98%D0%B7%D0%B4%D0%B0%D0%BD%D0%B8%D0%B5_%D0%92%D0%B8%D0%BA%D0%B8%D0%BF%D0%B5%D0%B4%D0%B8%D0%B8/%D0%9E%D1%86%D0%B5%D0%BD%D0%BA%D0%B0_%D1%81%D1%82%D0%B0%D1%82%D0%B5%D0%B9 [20:07:45] OMG UNICODE [20:09:34] cool, thanks. that's what Pratyush had already found, wasn't sure if there was a more canonical config location for groking that for ORES. [20:10:02] Regretfully not. Seems like an interesting idea though -- like a centralized reference. [20:10:07] I'm super excited about the progress on this Article Finder gsoc project. [20:10:31] linky? [20:10:37] such a thing exists for PageAssessments API, but ru.wiki seems to be the only one supported by ORES but not PageAssessments. [20:10:48] https://github.com/x-tools/xtools/blob/master/app/config/assessments.yml [20:11:12] ragesoss: PageAssessments extension is deployed on very few wikis. [20:11:22] There'd be plenty not supported by it. [20:11:34] Niharika: ORES wp10 is also for very few. [20:12:05] (I meant just wp10, in terms of 'supported by ORES but not PageAssessments') [20:12:15] Ah, okay, got it. [20:13:52] We're just about to release wp10 support euwiki :) [20:14:00] I imagine that doesn't have PageAssessment yet! [20:16:08] cool! euwiki have been pretty heavy dashboard users, so I think they might light what we're cooking up with article finder, once we get ORES hooked up for eu. [20:16:41] :D I'll be preparing a deployment today. We should be able to actually do it in Monday or Tuesday. [20:17:35] Nice. we'll probably also deploy multilingual support for the Article Finder next week-ish. [20:18:19] halfak: did you get a chance to try it out? https://dashboard.wikiedu.org/article_finder [20:37:47] halfak: It will depend on how long all of my interviews end up being, but if they are all an hour long, it will be about $670 to use a service. [21:05:29] ewhit_, so 5.5 hours of transcription? i think you should do it. It's gonna be a slog but it should only take one day. [21:05:38] I'll forward you an expense report form that you can fill out. [21:05:57] Oh wait one tick. Let me check to see if we have any licenses first [21:06:35] It's going to be about 9 hours of footage to transcribe. Yeah, it won't be fun, but it's doable! Thanks for checking on the licenses [21:07:47] Arg. yeah so two days. [21:07:51] And a headache [21:12:32] lol no worries, I'm a grad student, it's p much par for the course [21:38:57] ewhit_, no luck finding licenses, but I am getting more of a sense that we could just spend the $ on it. [21:39:33] The license is fairly cheap, and I have some discretionary funds from UM, so I can just buy one and get reimbursed. It would be a lot cheaper than the service [21:39:42] altho if the service $ is ok, I won't argue! [21:45:47] OK I can't find precedent for the service, so lets go with the software. [21:46:01] ewhit_, ^ [21:46:12] will do! [22:13:10] I'm out of here. Have a good weekend folks!