Posted by: Johan Normark | August 31, 2012

I am a man (79%)

Rosemary Joyce writes about a website (Genderanalyzer) that from the content of a blog analyzes the gender of the blogger (or is it the blog itself?). I made the test myself and they think that this blog is written by man (but they are only 79% sure). Is it because I write about climate change, Maya warfare, male-dominated object-oriented ontology, 2012, etc.? What makes them 21% not sure? My gender studies?

I entered the website’s own address to see if it can analyze itself. It did not work since an error occurred in the analysis: “Sorry, we could not extract enough text from http://genderanalyzer.com/. We could only find 178 words. We need more than 300 words to analyze a web page. More text gives better result!” Since the creators of the website have Swedish names I can make a 100% accurate analysis without the help of the website. Genderanalyzer is written by two men…

About these ads

Responses

  1. I tried the website that you referenced and it said that they were 75% sure that my blog was written by a woman. I hope in the future that it will be closer to 50%. I know there is supposed to be a focus for one’s blog, but I’d rather have fewer readers and be able to write on subjects that interest me, which tend to be eclectic. Hence the reason I’ve been reading your posts….

  2. Said they were only 71% sure I was man. No idea of the algorithm, but I’d guess they said male ’cause it’s science (yes, that’s sexist, but it is a fact that the majority of people in the hard sciences are men, though the demographics are very slowly shifting), but it wasn’t as sure ’cause I try to be non-antagonistic.

  3. Archaeology is less “hard” than astronomy from that perspective so it is a bit odd that I had 79%. I would also prefer to be around 50% as I try to be gender neutral. My whole perspective is to strive against anthropocentrism but that seems to make me androcentric according to this analysis.

    • I try to be gender-neutral, too. It’d be interesting to know what algorithm they use. Looking at how they did, though, it seems only to be a 52/48 split on right/wrong, so their algorithm is probably useless in its current state.


Categories

Follow

Get every new post delivered to your Inbox.

Join 128 other followers

%d bloggers like this: