The Computer Has Learned To Recognize The Gender Of A Person By The Peculiarities Of Using The Keyboard - Alternative View

The Computer Has Learned To Recognize The Gender Of A Person By The Peculiarities Of Using The Keyboard - Alternative View
The Computer Has Learned To Recognize The Gender Of A Person By The Peculiarities Of Using The Keyboard - Alternative View

Video: The Computer Has Learned To Recognize The Gender Of A Person By The Peculiarities Of Using The Keyboard - Alternative View

Video: The Computer Has Learned To Recognize The Gender Of A Person By The Peculiarities Of Using The Keyboard - Alternative View
Video: The Sentences Computers Can't Understand, But Humans Can 2024, May
Anonim

The nature of typing on the keyboard is different for everyone, but all men and all women have common characteristics that allow you to establish gender with confidence over 95 percent.

For their work, Ioannis Tsimperidis and his colleagues from the Greek University of Democritus selected 75 volunteers (36 men and 39 women). Specially created IRecU keylogging programs were installed on their computers, which recorded every keystroke for 10 months.

The collected files were analyzed by the ISqueezeU statistical analysis program. It allowed us to assess the possible correlations of a person's gender with different typing characteristics: the time of pressing the keys, the pauses between individual letters, etc.

The analysis showed that some of these characteristics do correlate significantly with gender. In an article published by the journal Digital Investigation, scientists mention the pause between pressing the N key and the following O, as well as M and O among such features.

These same characteristics and associated weights were used to train five different neural networks, which were soon able to determine the gender of a user with an accuracy of over 78 percent, with the most successful one exceeding 95 percent.

The authors believe that such an analysis can become a tool for discreetly and unobtrusively determining the gender of a user - for police investigations or planning targeted advertising campaigns.

Sergey Vasiliev