[17:16:49] <bmansurov>	 halfak: o/ What python library do you recommend for de-normalizing MW titles? Say I have "Early_Life" and I want to get "Early Life" from it.
[17:17:24] <bmansurov>	 I found out that mediawiki-utilities does normalization, but not the other way around.
[17:17:48] <halfak>	 http://pythonhosted.org/mwtypes/
[17:17:58] <halfak>	 Maybe that's where such a thing should live. 
[17:18:14] <halfak>	 Relevant: https://www.mediawiki.org/wiki/Mediawiki-utilities
[17:20:16] <bmansurov>	 halfak: thanks!
[17:42:34] <bmansurov>	 tizianop: o/ you there?
[17:42:48] <tizianop>	 bmansurov: yes
[17:43:16] <bmansurov>	 tizianop: which XML dumps are you using for generating the category network?
[17:43:55] <bmansurov>	 Here's a list for example: https://dumps.wikimedia.org/uzwiki/20180201/
[17:45:13] <bmansurov>	 tizianop: also, are you using https://github.com/attardi/wikiextractor to extract data from dumps?
[17:46:35] <dsaez>	 bmansurov: recomputing the recommendations now, you will have them in around 1:30 hour
[17:46:55] <bmansurov>	 dsaez: cool
[17:48:49] <tizianop>	 bmansurov: For the taxonomy provided, I'm not sure how the data is extracted. I sent an email to ask for more information
[17:49:35] <bmansurov>	 tizianop: great, thanks! I've updated the ehterpad with a JS library for denormalizing titles, no luck with Python yet.
[17:49:55] <bmansurov>	 tizianop: let me know when you hear back. Maybe the dumps already contain denormalized titles.
[17:50:56] <tizianop>	 bmansurov: I think it's better to use the SQL version. I found that some categories' parent are generated by a template that in the XML version is not expanded
[17:51:27] <bmansurov>	 tizianop: makes sense and in SQL titles are already normalized imo
[18:06:22] <Nettrom>	 bmansurov: If you have to write the Python code for denormalization, maybe have a look at pywikibot? It does handle titles, but not sure what the code looks like.
[18:07:03] <bmansurov>	 Nettrom: OK, thanks for the pointer!
[18:07:17] <Nettrom>	 yw :)
[18:07:21] <Nettrom>	 gotta run, back a bit later
[22:50:45] <leila>	 hey bmansurov. sooo, any update I should be aware of from the meeting with Tiziano et al? (sorry I just realized that you're probably out by now)