As weird as it may sound, but team of researchers at the Iowa State University (ISU) have developed algorithm to uniquely identify users based on their typing style and mouse usage patterns. These patterns are learned over a period of time and include the pauses user usually takes when typing words, searches for unusual characters, the spellings of the words that aren't very familiar etc. The algorithm can also analyze the patterns in usage of mouse & cursor to determine whether the user is indeed the authentic one. Research team says that they did a study on over 2000 users and the algorithm was correct about 99.5% of the times.


Professor of electrical & computer engineering, Morris Chang from ISU said that their technology is stable enough to distinguish between legit users and imposters. The technology operates behind the scenes and the user has no clue that he/she is being monitored. Professor Chang believes that the technology will be able to replace password with something that's more reliable and accurate in preventing data thefts or unwanted system logins.

The algorithm monitors user's typing rhythm for a period of 90 minutes and then stored in the central security system. The system checks its database whenever a login attempt is made to access sensitive data. If the system determines that a hacking attempt has been made, it can quickly log the user out of the system. It can be programmed to ask the user for a password.

This new system is likely to be used by DARPA so that the systems with sensitive information can be monitored throughout the time they are being accessed and not just at the time of login; as it currently happens.

Source: Iowa State engineers use keyboard, mouse and mobile device ‘fingerprints’ to protect data • News Service • Iowa State University