[17:28:54] halfak: is edit type identification something you've done yet? [17:29:07] yes. [17:29:17] Nothing deployed yet but it's on he agenda. [17:29:27] Probably a month unless there is a burning need. [17:29:45] Nothing deployed... in production? Or nothing deployed at all anywhere? I would be content using a beta service. [17:30:37] One thing I am interested in is using it to classify bots. Classifying bots based on their actual activity. [17:34:51] I've been experimenting with TF/IDF analysis on bot user pages. It's an interesting exercise but not without its challenges. I have a feeling the problems I am running into are similar problems that Google spends billions of dollars trying to fix [17:44:57] http://paws-public.wmflabs.org/paws-public/User:Harej_(WMF)/Lists_of_bots_on_every_wiki.ipynb?redirects=1 if you want to follow along at home [17:56:56] hi channel, qq: I have some questions about the Wikimedia API, do you know who can I contact? [17:59:17] !ask [17:59:33] eh, wrong channel for that, but i meant the bot to say how you should just ask [18:00:03] if somebody knows the answer they will respond. works better than trying to find who to ask [18:17:42] mutante: I want to use the parser on the Wikimedia API, because it takes all the templates, and works better than any offline version that I found, but given that I will do a lot queries, maybe is better to something else [18:18:28] dsaez: just throttle the number of queries you make and you will be fine [18:18:50] mutante, ok, that was my plan! thks [18:18:54] as long as you follow https://www.mediawiki.org/wiki/API:Etiquette [18:19:04] there shouldn't be a problem [18:19:31] maxlag parameter seems to be the key [18:19:48] yes, indeed [21:44:07] halfak: around? [21:44:31] o/ guillom [21:44:43] hey; may I PM quickly? [21:46:40] Sure :)