What is this temporary fashion or future technology?
There is such a thing in the world that 150 years already all scrapists and spyers used in their works - called "biometric authentication". It is different. The most common authentication methods are on the retina of the eye, on fingerprint.
Imagine: To enter your account on the Internet or to launch a blender, you need to enter the phrase "Mom Soap Frame". And so if you print it on the keyboard not as always - suddenly the finger cut or only asked? - Everything, your account is removed once and for all, and the blender begins to speak out!
Scary?
There is such a thing in the world that 150 years already all science and spyers used in their works - called "biometric authentication".
It is different. The most common authentication methods are on the retina of the eye, on fingerprint. Recently released "Agrusy man" We are in this a big assistant)
There is still a geometry of the hand (the brush is scanned with all its bends of fingers, thickness and length), geometry of the face, thermography of the face (do not believe - the signs of the image of blood vessels are studied!). And there are completely different methods - dynamic.
Dynamic methods include
- verification of signatures where special light feathers or electronic prints can be used, in many countries have legal force,
- And the analysis of the keyboard handwriting.
So here's the last speech and goes. The keyboard handwriting is not only the speed of input information, but also the intervals between the keystroke and the number of overlaps between them, the confusion time of the mouse, the degree of arrhythmicity when typing and using the function keys. It is even a frequency of errors when entering!
Everyone is sometimes written with errors (or oral eyes - how anyone like it) And everyone is guessing, where I learned Russian lettuce, and now I have an argument - this is my keyboard handwriting! Familiar to all stories that right hand Prints faster than left and therefore often the letters on the contrary turn out; And that I type in the pitch darkness and at night - when the light can not include and the head is already tired per day; that there was a bagel with boiled condensed milk and periodically swore on the keyboard; Barsik passed by ... Think - this is a lack of sleep affects, but no - the keyboard hand writing!
The analysis is now becoming in the priority of science, because people began to work remotely. Sitting under the palm tree, swinging in a hammock. With gadgets. Some prepare their written table to receive information - open the lid of the laptop, connect the devices and the stitching of the letter. And mostly everyone write on the knee, in traffic jam, in the subway, in line for Banderol in the post office of Russia ...
There is actually two ways to analyze the handwriting. The first - on the input of the famous phrase password and the introduction of an unknown generated by chance.
Any Khlebnikovsky "Ur Brees Shchur" is not counting here. Only adequate words. Moreover, that the system can learn all your physical capabilities, smelting and so on, you need to first learn to work with it - to find reference indicators. For reference: The diotable phrase password is longer - the better it will be, because when it is set, you will show yourself in all our glory.
But if you sit down for the device in the GM .. Not a very sober state, or with a stomach disorder, which will begin to suddenly manifest themselves in the form of (lowering the details) ... Consider on yourself. In this matter, even simple fatigue can play a role and will sit in front of the unfounded monitor until the morning.
Question: What if you have a laptop with a big keyboard, and the phone is small? You just need to configure the program for two typical input and NOU problems. And to write everything unambiguously and the first time - you can program a smart car. on automatic The beginning of reproduction of the same rhythmic music that will help you focus on entering information.
So, my dear, already autumn, and this is the best time to pass tests. Including keyboard.
Marsh, slender building!
Startup Typingdna has developed an extension for the Google Chrome browser, which recognizes a human keyboard handwriting. With it, you can create reliable two-factor authorization.
Keyboard handwriting
Under the handwriting, we usually understand the way to write, which each person has a special. Science Graphology is engaged in studying the handwriting - with its help, it is practically absolutely calculated, whose hand is written text. This knowledge is primarily in demand by criminalists, but sometimes the handwriting is interpreted to describe the psychology of writing and even his health.
More and more people go to purely printed text, so the graphology began to predict extinction. But the press has its own characteristics.
The keyboard handwriting takes into account the peculiarities of printing, and every person it is unique, as well as the usual handwriting. The keyboard handwriting is made up of input speed, input speakers, error frequencies and use keys to enter capital letters.
Typingdna
Typically, for two-factor identification, a pair of "login password" is used on the first stage, and on the second one more check: the control question, the code coming in SMS. Startup Typingdna decided to simplify the task and make the second step checking the keyboard handwriting. And it is checked until the user enters login with password.
The technique of definition of the handwriting is not the new technology itself, but Typingdna was the first to introduce it into the app. Previously, the startup released a utility in order to embed such checks on sites, but now it's still easier - just connect the page to the browser expansion. It will automatically determine the keyboard handwriting of the user when entering the password, and then offer to fill in a disposable request code.
Moscow State Technical University named after N. E. Bauman
Department of "Automated Information and Management Processing Systems"
abstract
in the discipline "Information Security"
"Keyboard handwriting as an authentication means"
Performed:
student of the group IU5-114
Pavlyuchenko K. G.
Checked:
CESEL S. A.
Moscow 2014
Introduction 3.
More about the method of 4
Testing method 8.
Conclusion 11.
Advantages of the method: 11
Disadvantages: 11.
List of references 13.
Introduction
In this abstract, consider one of the authentication methods - procedures for checking the authenticity of the access entity. The authenticator is some parameter provided by the system for verification.
Distinguish 3 types of authenticators:
Unique knowledge (password, pin-code)
Unique subject (key, smart card)
The unique characteristic of the subject itself (static - fingerprints, snapshot of the retina eye behavior - for example, voice authentication)
The latter type also includes keyboard handwriting.
This is a behavioral biometric characteristic that the following parameters describe:
Input speed - the number of characters entered, divided by the printed time
Input dynamics - is characterized by time between keystrokes and the time of their retention
Frequency of errors when entering
Use keys - for example, what function keys are pressed for the input of capital letters.
More about the method
All people perceive the events that occur in different ways. Try for a short time Ask the number of points or vowels in long words, the size of horizontal and vertical lines, - how many subjects will be, so many opinions. These features of human psyche are also suitable for identification. True, depending on the state and well-being of a person, the obtained values \u200b\u200bwill "swim", so in practice they rely on an integral approach when the result is supplied to several checks, given and work with the keyboard. For example, a method of identification may be: on the screen, for a few seconds, vertical or horizontal lines appear. Their size and number is random. The user dials the corresponding, in his opinion, the numbers. Thus, it turns out: the characteristics of the keyboard handwriting are evaluated how the indicated length and the number of lines are close to reality, attention and accuracy of counting (as far as the same line is properly compared with the neighboring). And finally, the results are compared with the standard. In this method, errors in the sizes are not so important, the main thing is that they are repeated and when setting up, and when identifying.
From the point of view of using hidden monitoring of computer security systems, the classification of the psychophysical parameters of the user, to which they include: keyboard handwriting, mouse signature, reaction to events occurring on the screen. We will dwell only on considering the use of the keyboard handwriting to identify the personality.
One of the sufficiently complex tasks everyday solved by many people is a quick set of texts from a computer keyboard. Usually, fast keyboard input information can be achieved through the use of all fingers of both hands. At the same time, each person manifests itself its unique keyboard handwriting. The keyboard handwriting is a set of dynamic characteristics of the keyboard.
Not many guess that in communication with the computer, the individuality of the user manifests itself in the speed of the character set, the habits use the main or additional part of the keyboard, the character of the "dual" and "built" keypads, in favorite computer management techniques, etc. Nothing amazing is akin to Melomanan's ability to distinguish between pianists, performing the same work, or the work of telegraphists who use the Morse code.
This identification method is popular in the United States to prevent children in the Internet through home computers. Even if the child spied or recognized the password of his parents, he could not use it. Also, this method can be used for additional protection when organizing access in computer systems.
The identification of the keyboard handwriting is to select the appropriate standard from the list of stored standards stored in the memory of the computer, based on the evaluation of the proceedings of this reference parameters of the handwriting of one of the operators entitled to work with this computer. The solution of the user identification task is reduced to solving the image recognition task.
A classic statistical approach to the user recognition on the keyboard handwriting (a set of keywords) revealed a number of interesting features: The dependence of the handwriting from alphabetic combinations in the word, the existence of deep bonds between the set of individual characters, the presence of "delays" when entering characters.
A highly important characteristic of biometric identification is the length of the password phrase. Practice shows that the password phrase should be easily memorable and contain from 21 to 42 clicks on the keys. With the synthesis of the password phrase, it is permissible to use words with meaning.
In addition, there may be an analysis of such signs as the dependence of the input rate of words from their meaning, the relative time of pressing various keys, etc., and they are even more informative: for example, the reaction to the various terms will indicate the scope of its interests. Indeed, the chemist is faster to "hydrogen", "compound" than the "program" or "excavator". And the model will be familiar with such words as "mannequin" or "pattern".
Collecting biometric information about the user's operation When analyzing the keyboard handwriting occurs with the measurements of the intervals between keystrokes and the time of their retention, after which the results obtained are formed in the intersomymbol interval matrix and the keys retaining time vector. After collecting biometric information, the data obtained is compared with their reference values.
How can the individual features of the keyboard handwriting be identified? Yes, as well as in the Graphological Examination: We need a reference and studied text samples. It is better if their content is the same - the so-called password or key phrase. Of course, in two-three, even ten pressed the user, it is impossible to distinguish the user. Need statistics.
When typing a key phrase, the computer allows you to fix many different parameters, but to identify most conveniently use the time spent on entering individual characters. TL, T2 key press times, ... TN is different and, accordingly, the values \u200b\u200bof these parameters can be used to identify the characteristic features of the keyboard handwriting of the user. In addition, you can use as monitored parameters of the intervals between pressing adjacent keys. Controlled parameters are significantly dependent on how many fingers use the user when typing, from characteristic combinations of movements of various fingers and from the characteristic movements of the hands when set. For example, if you force the user to work in one finger of one hand, the keyboard handwriting almost completely loses its individuality. In this case, the time of pressing the keys cease to reflect the individuality of the people, since the intervals between the presses become proportional to the distance between the keys, and the overlapping of the adjacent keys becomes impossible.
The unique features of the keyboard handwriting are detected by two methods:
on a set of key phrase;
on the set of "free" text.
Each of them necessarily has configuration and identification modes. When setting, the reference characteristics of the user's input key phrase, for example, the time spent on individual characters are defined and remembered. And in identification mode, the reference and obtained sets are compared after the exclusion of gross errors.
A set of "free" text is performed according to the most diverse phrases (the key phrase, as a rule, is the same), which has its advantages, allowing you to obtain individual characteristics of imperceptibly, without focusing the attention of the user on the password phrase.
The selection of the scan schema depends on the application with which it is used. For example, if an accountant wanted to get a short help, and the computer instead proposes to gain 2-3 page "free" text to make sure that it is really the right person. There are not enough nerves and it will only cause irritation, and as a result - the user will try to avoid such an identification system in every way.
On the other hand, the one who has tolerance to the secrets can work with such a program all day, from time to time it is from the computer. And so that at this moment the attackers did not take advantage of the opened system, it is advisable to periodically perform a "unlaspped check". Such systems allow you to constantly monitor whether the user is legitimately sitting at a computer.
It should be noted that when using these methods, it appears not only to confirm the authenticity, but also to analyze its condition. The described approach to protection against unauthorized access allows:
control the physical state employees;
end the practice of violation of security rules when working with passwords;
provide a simpler and the same reliable login method.
Authentication - Procedure for checking the authenticity of the subject. Authenticator - Some parameter provided by the system for checking.
Distinguish 3 types of authenticators:
- Unique knowledge (password, pin code)
- Unique thing (key, smart card, token)
- Unique characteristicsubject (static - fingerprints, retina snapshot, behavioral - for example, voice authentication)
The last type also applies keyboard handwriting.
Name \u003d "More"\u003eKeyboard handwriting - The behavioral biometric characteristic that the following parameters describe:
- Speed \u200b\u200binput - The number of characters entered divided by the printed time
- Dynamics of input - characterized by time between keystrokes and the time of their retention
- Frequency of errors when entering
- Using keys - For example, what function keys are pressed for the input of capital letters.
To test authentication capabilities, using the keyboard handwriting, we used a simple self-written program. You will come up with some password phrase, which will make it possible to repeatedly enter the subjects (in our case, the password phrase is "orange").
The rate of input for one evening will vary in some peys. Using the statistics received, we will get some mat. Waiting for speed and its dispersion.
After conducting similar tests in other psychophysical states, we will get mat. Lamination and dispersion and for them.
The figure below shows the status:
- first column - evening (only wrote the program)
- the second column is morning (only woke up)
- third Stage - Another evening
- fourth Column - Other Keyboard
Now it is necessary to compare the results of speed tests with other subjects.
Despite the differences in the printed speed of all subjects, there is a speed range in which they all could print the password phrase (drawing on the right). Therefore, in addition to the speed to reduce the number of false positives, it will be necessary to use other characteristics of the keyboard handwriting, or rather the dynamics of the password phrase.
In the picture on the right, the height of rectangles describes the time intervals between the presses of the adjacent characters in the Orange Password phrase.
Compare with other states:
Using another keyboard and entering the password phrase in the state "just woke up" leads to significant changes in the dynamics of input, so to reduce false failures, the use of the same keyboard is required, and authentication is carried out with regard to day time when it occurs.
Now compare the input dynamics of different subjects:
We will receive the imposition of graphics input dynamics on each other, again, that there is a probability of coincidence of the dynamics of various subjects, but this probability is already significantly lower than when accounting only input speeds.
Advantages of using the keyboard handwriting for authentication:
- Easy to implement and implement. Implementing only software, the input is carried out from a standard input device (keyboard), which means that the use does not need to purchase any additional equipment. This is the cheapest way to authenticate the biometric characteristics of the access entity.
- Does not require any additional actions from the user except usual. The user is one way or another, for sure, uses a password that can be assigned to a password phrase, which will be authenticated.
- The possibility of hidden authentication - the user may even be noted, which includes an additional check, which means it will not be able to inform the attacker about it ...
Disadvantages:
- Application learning required
- A strong dependence on the ergonomic keyboard (in case of shift, you will have to train the program again)
- Strong dependence on the psychophysical state of the operator. If a person fell ill, then it is likely not to be authenticated (on the other hand, it may not be necessary to do in a sick state).
We are engaged in consumer lending online. The issue of preventing fraud for us is one of the main. Often, potential scammers think that the security systems of the Fintech companies are lower than in the traditional financial organizations. But this is a myth. Everything is exactly the opposite.
In order to reduce the level of potential fraud. We developed a behavioral biometric system that is capable of identifying a person on the keyboard handwriting. The pilot version was introduced in Russia in August 2017.
How it works
The system allows you to authenticate the user on behavioral patterns inherent to each person. Among them: the speed and dynamics of the text set, the transition time between the keys, typos and the features of the mouse cursor movement on the screen. As soon as a person is registered in the Personal Account, the system is analyzing his behavior and makes the cast, which is used when subsequent authentication on the site.New program Excludes cases of fraud. If the user's handwriting does not match the lock in the database when entering the personal account in the database, this profile is entered into a list of risky requiring additional check. Those. If the system has reason to assume that a person is not the one for whom it gives out, it can request additional data or send this appeal to verification, where a living person will already decide to approve or refuse a loan.
The peculiarity of the system is that the analyzed behavioral patterns cannot be faked: even if a person changed the keyboard, they can change slightly, but in the aggregate, their structure will remain unchanged.
