Browse Prior Art Database

User Validation Based on Typing Characteristics Disclosure Number: IPCOM000247519D
Publication Date: 2016-Sep-14
Document File: 3 page(s) / 32K

Publishing Venue

The Prior Art Database


Passwords are used almost always used to validate and protect the access to systems/devices. External devices, such as passwords generator dongles, biometric parameters scanners (like finger print or eye scanners) can be used to make a more safe user identification. But these methods, besides software, require additional hardware and additional costs. Hereafter it is disclosed a system based on software which makes more safe the user identification, based on his typing characteristics.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 39% of the total text.

Page 01 of 3

User Validation Based on Typing Characteristics

    Usually passwords are used to validate and protect the access to systems/devices. There are systems were the password is continuously generated by a portable dongle device which is in synch with a centralized system, but this does not guarantee that the user inputting the password is the registered user and the one entitled to do it. A way to identify the user who is inputting a password, is based on the use of biomedical information such as finger prints. This will provide a reliable user identification, but, besides SW, it requires additional HW devices (such as finger print readers) attached to the keyboard input system.

    Hereafter, it is disclosed a system and a method to verify the user identity and increase the identification accuracy, also basing on the way he normally types each of the keys (frequency, speed, typo errors etc). Basically the system learns the "typing" characteristics of the user, and it is able to recognize an user, within a certain percentage of reliability. The characteristics used are based on the typing speed, on the typing of specific sequences of letters, the time between one letter and another, the difference between left side and right side keyboard letters, common and repetitive typing errors (like letters switching) and so on. This system can be used in conjunction with a normal password and can provide a means to try identifying the user typing the password or pass phrase (a standard sentence which can be proposed to be typed and can be compared with the standard one for that specific user).

    The main new block components of the proposed system are shown in Fig. 1 and are listed hereafter :

- Typing metrics evaluation logic
- Self learning logic based on selected metrics
- Checking/validation logic

Figure. 1 Component blocks of the identification system


Page 02 of 3

Typing Metrics Evaluation Logic (new)

    This component evaluates the user typing metrics based on the info it is requested to detect. Some of these metrics (not limited to), can be:

- Typing rate. The system detects the average speed the characters are typed on the keyboard. The time between one character key and the following typed is evaluated.

- Left/Right keyboard side use. The typing rate breakdown between the left keyboard side positioned keys vs the right keyboard side positioned keys is evaluated.

- Close/far keys use. The typing rate breakdown between close and far keyboard keys is evaluated.

- Typical typing errors. One typical error for example might related to letters switching

- Specific keys use. For example the use of the left vs right shift, caps lock vs shift for upper case, num lock key vs number keys, Backspace vs Del to delete, Ins vs No Ins, PF keys.

- Positioning. Arrows keys vs mouse click or tab vs enter key.

- Other characteristics might be related to the use of a specific language, keyboard layout or keys remapping.

- Specific patterns identification (sequence of l...