[15:31:59] o/ [16:09:57] ciao leila [16:10:01] good morning :) [16:10:12] miriam_: ciao. :) [16:10:15] what's up? :) [16:11:13] leila: I am going to a Wikidata meetup tonight, right after our meeting, so I will be moving "office" shortly to get closer to the venue! [16:11:38] miriam_: sounds good. [16:11:43] miriam_: what Wikidata meetup? [16:12:15] leila: it's organized by john from Wikimedia UK, he wants to do a monthly gathering of wikidata enthousiasts [16:12:22] kicking off today [16:13:37] miriam_: that's nice. [16:14:08] leila: yes, I am happy to meet the local community! [16:27:36] miriam_: what is the work on coverage of Alexa web categories? [16:27:46] (reading through the notes of last week) [16:30:27] leila: we annotated all external references with alexa web service categories, for the citation usage project [16:30:48] we have 35% coverage only, because AWIS doesn't categorize all websites [16:31:14] so we'll need to annotate more in crowdsourcing, if we have funds for that :) [16:31:31] miriam_: oooo. now I understand. thanks. [16:38:34] exciting [16:38:39] what are we going to do with all this labeled data :D [17:02:55] harej: understanding better why readers interact with citations! [19:40:38] Hi research people :) [19:40:48] hey joal ! [19:40:56] dsaez: would you be nearby? [19:41:08] Hi isaacj - I have seen your comment [19:42:04] isaacj: I'm glad the data is usefull, I also want to make sure you keep a critical eye when playing with it - There might be some gotchas not foreseen (room for improvement !) [19:42:58] yeah, i could see a use-case where you want to know what the interlanguage links were at the time for a revision but that would be different it seems (and probably a lot more painful to generate) from the revision table you generated [19:43:25] joal: sounds good. i'll make sure to do some spotchecking as I go and let you know if anything comes up [19:43:41] correct isaacj - For this type of data we'd need to parse wikidata-json-dumps and extract temporal links (very doable though) [19:44:38] i guess main challenge would be how to handle revisions before the switch to wikidata [19:45:02] but not my use-case so i shouldn't conjecture too far :) [19:45:07] isaacj: I don't even want to think of that :) [19:45:49] isaacj: This exact use-case you're describing is the type of things I'd like to be able to generate from having the dumps on hadoop - But not now :) [19:47:06] joal: yay! [19:48:19] isaacj: I think halfak might recall a time when youngly hired joal was willing to extract everything from text :) And here we are, 4 years after, and text is just at its beginning :) [19:49:28] hahaha [20:07:54] joal, :P [20:08:20] It's hard work [20:09:07] It is, halfak, and not at the top of the list, so it waits