[04:10:57] Hi I am Stefanie Muroya! I find this project very interesting and I would like to take the opportunity as an Outreachy applicant to contribute to this project and learn the most I can! [04:11:34] "A system for releasing data dumps from a classifier detecting unsourced sentences in Wikipedia" project ***** [14:16:01] @Samwalton9 hello...I am interested in learning more about Machine learning project for Outreachy internships [14:17:44] Samwalton9 If you have any specific instructions as to how to approach and start contributing it would be very helpful for me [14:28:19] Hi Shamima_19! We've actually just been talking about this because we'd like to create some better newcomer tasks. For now though, we have an onboarding task at https://phabricator.wikimedia.org/T233709 which we recommend you follow to make a start with using the model :) [14:28:49] Unfortunately we don't really have a tool or code that you can already contribute to, because this is a new project. [14:56:21] djellel: congrats on your IRC handle. welcome here, too! [14:58:08] Thanks! Hello IRC, it's been a while :) [15:02:03] does anyone have a paper they like when it comes to representing Wikipedia articles based on their links? for instance, finding related articles based on how many links they share [15:02:11] (and welcome djellel !) [15:06:05] Hi isaacj. You mean using internal page rank? [15:06:34] djellel: one question for your intro email. Is this a good place to link to? https://dedcode.github.io/ (I default to google scholar otherwise) [15:07:07] djellel: not so much a metric of importance for a given article but a metric of similarities between articles based on their shared links [15:08:12] so trivially you can to tf-idf on an article's text and then find articles with similar tf-idf but you could also do something analogous that had to do with graph of wikipedia articles (based on wikilinks) and i'm looking for a good reference paper for that [15:10:04] isaacj: if you don't find an answer here, ask Bob. [15:10:09] leila: Yes, the github page is fine [15:10:15] djellel: thanks! [15:10:39] leila: good point -- thanks [15:12:10] isaacj: Random walks [15:12:21] Samwalton9 this is weird but I just refreshed the page and can't seem to find your previous replies to me..is this how the site works? what is the way to save messages pls? [15:12:57] Ah, yes, an unfortunate result of the way IRC works. [15:13:05] I can paste my messages again if that would help [15:13:06] djellel: yes, that would be a likely strategy. i'm looking for a paper that has looked into this and compares approaches [15:13:20] isaacj: a quick search yields this https://nlp.stanford.edu/pubs/wikiwalk-textgraphs09.pdf [15:14:19] djellel: thanks! [15:15:16] isaacj: You can probably expand your search by looking up "entities" [15:16:04] * leila heads out to grab a breakfast and start the day in WikiLead. See you later in the evening PST or tomorrow when I'll be back to usual work hours. [15:16:47] Samwalton9 would be wonderful if you could do so as I lost them. Thanks a lot! [15:17:14] Hi Shamima_19! We've actually just been talking about this because we'd like to create some better newcomer tasks. For now though, we have an onboarding task at https://phabricator.wikimedia.org/T233709 which we recommend you follow to make a start with using the model :) [15:17:16] Unfortunately we don't really have a tool or code that you can already contribute to, because this is a new project. [15:17:32] We're going to create one or two more starter tasks this week. [15:26:22] Samwalton9 Should I for the task to be assigned to me ? Like in github. Or is this a general task for anyone interested to work in this project. Sorry this phabricator thing is new to me . [15:26:55] It's a general task, no need to assign it to yourself. You might want to sign up/log in to Phabricator so you can post questions on that task though [20:03:44] Hello, I am Busirah Hammed. I am outreachy applicant, I am interested in Wikimedia-research. [20:05:51] welcome, h_bushro! [22:41:06] Maybe the topic should have some message for outreachy applicants [22:41:15] (I don't think it's obvious to them that they won't necessarily get a response right away) [23:18:36] Hello everyone, this is aya from egypt. I'm one of the applicants for the outreachy internship [23:19:46] hope this is the right place to make this note, the phabricator link posted with the project descriptions returns a 500 internal server error [23:20:11] this one: https://phabricator.wikimedia.org/T233709 [23:20:34] and also this: https://phabricator.wikimedia.org/T233707 [23:20:57] aya-s: they're working for me [23:21:22] hmm, I am logged in, though [23:21:48] seems to work logged out too [23:23:23] oh that's very strange :S, I'm actually unable to open https://phabricator.wikimedia.org [23:23:25] as well [23:23:36] same error, 500 [23:25:25] are you using a proxy or VPN of some sort? [23:27:00] I'm not, I'm able to access a cached version from 28 september [23:27:23] I will check to see if it is an ISP issue of some sorts maybe