Are You a Male, or Are You a Female?
Although internet is free for all, it’s not a secret that some websites is written and visited by mostly one gender. Tech and gaming site is mostly written by male and women’s fashion magazine will be populated mostly by female. Not that women in tech site or men in fashion blog is bad, the preference of certain gender in visiting a website is of course useful for advertiser.
By using word algorithm and training, by the writing only, an author’s gender can be guessed by at minimum 300-word writing. It is based from a research from a team of Illinois Institute of Technology and Bar-Ilan University in Israel (Shlomo Argamon, Moshe Koppel, Jonathan Fine, and Anat Rachel Shimoni) which utilizing weighted word frequencies and parts of speech by Bayesian network. Because the research is based on training and probability, it will not be 100% accurate, it was written that the method will be 60-70% accurate. With 50% odd of random guessing, that much is good. (There are some people who really feel suspicious with it. It is basically training an artificial intelligence. Give an example and showing it the right answer, rinse and repeat until the number is adequate. Then test it to other controlled writings and calculate the percentage of result.)
Estimation is based mostly by word preferences, and there is difference between fiction and non fiction and also formal and informal writing. There are many factors however influences someone’s writing. Field of work, presence of editor, knowledge, age, experience, these are some of them.
For analyzing website, use GenderAnalyzer
Other than male or female guessed by writing, you can also have your age guessed by your photograph here.
Have you tried to put your writing inside the analyzer? How does your writing analyzed?
Note: It seems that this blog is relatively gender-neutral. It is analyzed as 51% written by female. I guess it is because of this blog confusing identity as half-tech blog and half-culture blog.