Re: [Gendergap] Computational Linguistics Reveals How Wikipedia Articles Are Biased Against Women

2015-02-04 Thread WereSpielChequers
I think this one is worth looking beyond the headline.

There are two specific areas where we fail, in the language we use when we 
write about women and in the relative lack of links to articles on women.

The two areas where the study indicates we are doing OK are the two where we 
have put in a lot of work over recent years, covering the men and women in the 
same ratio as those benchmark sites, and putting women on the main page. Of 
course those areas are only OK if we accept that our task as a tertiary source 
is to reflect but not magnify the skews in the secondary sources.

it would be good to know if the relative paucity of links to articles on women 
was simply down to fewer of the mentions of women being linked, or we had a 
deeper problem in that women were less likely to be mentioned in other 
articles. One problem is rather easier to fix than another, as a community we 
have been looking for new entry level tasks for some time, and adding more 
links to underlined articles could easily be one of them. Especially if we can 
get lists of articles with few incoming links but multiple other articles that 
appear to mention the subject. I think I'll file a bot request for that. 

Regards

Jonathan/WereSpielChequers


 On 2 Feb 2015, at 22:12, Rob gamali...@gmail.com wrote:
 
 MIT Technology Review: Despite well-publicized efforts to promote
 equality, Wikipedia articles are deeply biased against women, say
 computer scientists who have analysed six different language versions
 of the online encyclopedia.
 
 http://www.technologyreview.com/view/534616/computational-linguistics-reveals-how-wikipedia-articles-are-biased-against-women/
 
 ___
 Gendergap mailing list
 Gendergap@lists.wikimedia.org
 To manage your subscription preferences, including unsubscribing, please 
 visit:
 https://lists.wikimedia.org/mailman/listinfo/gendergap

___
Gendergap mailing list
Gendergap@lists.wikimedia.org
To manage your subscription preferences, including unsubscribing, please visit:
https://lists.wikimedia.org/mailman/listinfo/gendergap

Re: [Gendergap] Computational Linguistics Reveals How Wikipedia Articles Are Biased Against Women

2015-02-04 Thread Sydney Poore
Hi, I think that there is a problem of not creating articles about women
who are notable in there own right but are mentioned in articles of family
members.

It is not uncommon to do use Wikipedias search for a notable woman and find
a mention of their name in articles about a family member. It can happen
the other way too, but I think there is still more of a a bias towards
thinking woman who are notable are daughters or wives of notable men and
looking for that link.

So, I encourage everyone to look for the link both ways equally.

Sydney
On Feb 4, 2015 7:07 PM, WereSpielChequers werespielchequ...@gmail.com
wrote:

 I think this one is worth looking beyond the headline.

 There are two specific areas where we fail, in the language we use when we
 write about women and in the relative lack of links to articles on women.

 The two areas where the study indicates we are doing OK are the two where
 we have put in a lot of work over recent years, covering the men and women
 in the same ratio as those benchmark sites, and putting women on the main
 page. Of course those areas are only OK if we accept that our task as a
 tertiary source is to reflect but not magnify the skews in the secondary
 sources.

 it would be good to know if the relative paucity of links to articles on
 women was simply down to fewer of the mentions of women being linked, or we
 had a deeper problem in that women were less likely to be mentioned in
 other articles. One problem is rather easier to fix than another, as a
 community we have been looking for new entry level tasks for some time,
 and adding more links to underlined articles could easily be one of them.
 Especially if we can get lists of articles with few incoming links but
 multiple other articles that appear to mention the subject. I think I'll
 file a bot request for that.

 Regards

 Jonathan/WereSpielChequers


  On 2 Feb 2015, at 22:12, Rob gamali...@gmail.com wrote:
 
  MIT Technology Review: Despite well-publicized efforts to promote
  equality, Wikipedia articles are deeply biased against women, say
  computer scientists who have analysed six different language versions
  of the online encyclopedia.
 
 
 http://www.technologyreview.com/view/534616/computational-linguistics-reveals-how-wikipedia-articles-are-biased-against-women/
 
  ___
  Gendergap mailing list
  Gendergap@lists.wikimedia.org
  To manage your subscription preferences, including unsubscribing, please
 visit:
  https://lists.wikimedia.org/mailman/listinfo/gendergap

 ___
 Gendergap mailing list
 Gendergap@lists.wikimedia.org
 To manage your subscription preferences, including unsubscribing, please
 visit:
 https://lists.wikimedia.org/mailman/listinfo/gendergap
___
Gendergap mailing list
Gendergap@lists.wikimedia.org
To manage your subscription preferences, including unsubscribing, please visit:
https://lists.wikimedia.org/mailman/listinfo/gendergap