How we perceive our world is surely influenced by what kind of people we let represent it. Lead roles in movies is mainly held by men (at least in the top #100 imdb movies). That sets the stage for a world where men are perceived as in the lead role. This is perhaps not news, but quantifying this inequality might be a step in the direction of correcting it.

Therefore I built a bot that:

1. Daily scans the major Swedish news sites for names of the form Firstname Surname.
2. Checks every firstname against Sweden Statistics (SCB) name database, counting every name used by more than 95 % of exclusivly one gender. Unisex names are therefore excluded.
3. Calculates the percentage of women represented by each site and logs it in a timeseries.
4. Tweets once every week and month with a graph ranking the Swedish news sites.

It looks something like this.

So basically it shames the least equal newspaper, hoping for retweets to add on to the feel of a angry online mob. In this analogy, twitter is the public shame pole of the village.

The timeseries shows some general trends already, even though the bot was broken for some time (as you can see from the graph).

Etc.se which is a left wing newspaper works activily with equal representation in their texts and it obviously reflects on the data. Most other newspapers lies in the span 20 % - 35 % women represented. Svt.se which is the Swedish public service, that is funded by tax money (black line), comes out second.

When more data is avalible I will probably revisit this again.

Code for the bot can be found on my Github and the twitter account connected with it is Shamebot3 (shamebot one and two was already taken).