Posts: 48
Threads: 18
Joined: Oct 2009
Reputation:
0
[attachment=6704]
ABSTRACT
The deficiencies of traditional password-based access systems have become more acute as these systems have grown in size and scope. Researchers are actively investigating ways to improve the security of password systems or offer replacements. One category of improvements uses keystroke biometrics, an approach which seeks to identify an individual by their typing characteristics. Since 1980, a number of techniques have been proposed for accurately harnessing keystroke dynamics for system authentication and other novel uses. But do these systems deliver on their promise to increase system security and simultaneously ease the burden of logging into systems and remembering passwords? And do databases of users' keystroke profiles present additional privacy concerns?
The keystroke bio metrics is used with the application in news reporting system. It will detect the person who send the news is the reporter or some other person who hacked the user name and password of the system. First the pattern of the reporter is stored with the server system. Server after receiving the text then it matches with the text pattern information in it.
INTRODUCTION
As an attempt to develop more powerful authentication system, with low cost and good acceptance by users, we proposed here an authentication mechanism based on biometric information of human typing patterns. At the movement we present a complete description of the architecture with the multi pattern verification units and results of first phase implementation of the mechanism based on single pattern verification unit among Multilayer perceptron, Learning-Vector-Quantization, Self-Organizing Neural Networks and Support Vector machine.
Today, all computer based system claims for more sophisticated mechanisms to guarantee the information security. The fast evolution of communication systems provided us a great volume of information anywhere any time. The security question became proprietary. Making these systems reliable and secure is one of the most important challenges of the communication evolution.
Authentication is the way to correctly verify one person who he or she claims to be. Many research works have been developed on the way to correctly identify somebody. Since old times, the humans try to identify each other correctly. The most traditional way to confirm that some body who he or she claims to be is to verify his or her handwritten signature. In computer systems, similar issue should be considered.
All information systems adopt some king of authentication. The most common mechanism is called user name and password. This mechanism consists of basically of an association between a public information (username normally every one knows it) that uniquely identify the user on the system, and a secret word (no body beyond the user should know it) that confirms that the person associated by that user name is who he or she to be. This mechanism presents some drawbacks that make it very week. Some of its drawbacks are: Persons choose easy to break passwords as family names and birthday dates; persons normally write their passwords in places of easy access; one can easily see one password on steal it with no knowledge of disclosure. On the other hand, to develop an additional authentication mechanism that carries some advantages like low cost; high performance and high acceptability are not easy. User name and password are easy to implement (low cost) and are largely acceptable by the users. That s why still most famous authentication mechanism applied nowadays.
There are three main techniques to verify ones identification: some thing a person knows (a code); some thing a person posses(a card); some ting a person has(a characteristic). All these three techniques can be combined on the way to produce more efficient identification system. Naturally if we apply all the three techniques together a more secure authentication mechanism will be produced. However we still have to evaluate the cost and the acceptance issues involved in establishing a more sophisticated authentication mechanism.
The last technique is based on one s biometrics characteristics. A biometrical system is a pattern recognition system that establishes the authenticity either specific physiological characteristics (some particular structural characteristics such as hand size or iris format and color) or behavioral characteristics (some particular behavioral characteristic such as typing speed or writing pressure) inherent to a user. .
One kind of biometrical behavioral characteristic that can be used to provide a particular identification is the dynamics characteristics of someone s typing or the human typing pattern. Many studies shown that this approach is possible and effective.
Combined with traditional authentication system, the user typing information can be of some help to identify users more precisely. Considering behavioral typing information to authenticate users can be very convenient because not extra hardware is necessary. All the behavioral information can be obtained by software systems, what generally implies lower cost than hardware development. Moreover nothing changes in the way the user authenticates himself, what makes it more acceptable. However the great question that lies on the kind of authentication is how precisely can we verify the user typing behavioral characteristics. Many studies, as mentioned before, have shown that this kind of identification is viable but still far from obtaining satisfactory indices of FAR (False Alarm Rate) and IPR (Impostor pass rate) if compared with other biometric techniques like finger printing. We will show that FAR and IPR go on opposite directions It means if we try to reduce one of these indices the other will grow up.
FAR False Acceptance Rate- indicates how frequently the system rejects the valid users. IPR- Imposer Pass Rate Indicates how often systems accept an imposter as an valid one. Those indices are most common metrics to evaluate the performance of biometric systems on their capacity to make mistakes on a matching pattern.
To evaluate typing characteristics two main measures have to be obtained about ones typing:
Inter key time- The time between two consecutive key releases and press.
Press time- The time between one key press event and key release event.
In our solution we will use a combination Inter key time and press time for effective results.
Lets consider two keys x and a are pressed. Let us consider Ta is the press time for key a. Tx is the press time for x. and Tax is the inter key time between Ta release event and tx press time.
This measure can be obtained through typing data collector that read keyboard events and processes this data for feature analysis.
As an attempt to develop good authentication system we proposed here a mechanism that combines two techniques of identification: something that the person knows( i.e., a password) and some thing that a person has (typing characteristics). In this work we are considering static authentication or authentication performed during the access to the system.
Results presented in this work indicate that combining two or more techniques can bring better results in terms of IPR and FAR. In statistical approach FAR and IPR Decrease 41% and 22% respectively in neural network approach 21% and 6% respectively.
The basic idea behind the model proposed here is the support some results presented by Brown and Roogers, which show that some techniques used successfully to verify one s typing pattern could present no good result for other person. It means for some users some techniques would give best results than others.
So an authentication mechanism working with a large group of users to be authenticated should obtain best global results in terms of FAR and IPR if it combines different techniques of typing pattern verification, and uses the same technique for each user.
Our purpose is to implement a mechanism of authentication with 3 techniques and combine it with a cluster analysis to segment users in 3 groups. Each group will be analyzed by one pattern matching technique that should give best results for this group of users.
The authentication mechanism proposed here can operate in two modes: new user registration and user authentication. In the first case, the mechanism will record the user username, password and typing profile.
The typing profile is then analyzed and stored so that it can be used during the authentication phase. In this mode, the user will be asked to type his user name and password about 10 times.
On user verification mode the system will verify if user name and password are correct. After authentication has been done, the typing characteristics of that user are compared to his/her stored profile. If it s typing pattern is quite similar to the stored profile the user is granted access. So the user identifies him self or her self by a user name, a password and the typing characteristics.
The authentication mechanism has five main bocks: registration, password verification, identification, typing verification and decision.
Registration Block: This modules is responsible for registering the user in the system, It begins by obtaining user data {UN- user name, PW password, RS-10 Typing Reference Samples}. After that, a cluster analysis is run to identify the best group for the new user. After the definition of the cluster, each verification unit should be prepared to evaluate that user pattern. Although we are defining a cluster to choose which technique should be used for each user, we will prepare all 3 verification units to analyze any user pattern. Our intention is to keep the operating a multiple verification system and compare its performance with our best verification technique in a future study. For each verification unit, after the training sessions, the registration block will produce a configuration matrix (M) for each verification unit to be able to do the authentication of specific user. In this way, after registration, each user will have on profile database: the username(UN), password(PW) and 3 matrix of configuration (M)- One for each verification unit- and cluster © he or she belongs to.
Password Verification: during the authentication process this block is responsible for searching the username(UN) and password (PW) information on the profile database and for verifying if the username and password combination is correct. If it s not, the user is denied access, other wise the user goes to the next phase of authentication: typing verification.
Identification Block: After a user has been correctly authenticated by password, this block searches for the configuration matrix for this user and the cluster to which him/her each belongs to. So it feeds the next block with the configurations matrix and test typing sample to be verified. If we are working in multiple verification mode, this block will send to next the configuration matrix for each verifier unit. Other wise if we are operating in best verification mode, only the configuration matrix of the verification unit associated to the cluster the user belongs to will be sent.
Verification Block: This block processes the test sample sent by the identification block. To correctly process the sample this block should configure each verifier unit with its specific matrix and so test the sample.
The decision block is responsible for deciding if the result presented by the verification unit makes that test sample valid or not for that user. Generally it will decide based on the threshold that will represent the sensibility of global verifier. If this threshold is high, or make easy to pass through the system, its expected to have low FAR and high IPR. But if not its expected to have high FAR and low IPR. That s big tradeoff of any biometric system; if FAR increases IPR decreases.
If more than one unit is used to process the test sample, the decision block should analyze all results and decide in some way if it is or not a valid sample.
Posts: 69
Threads: 17
Joined: Oct 2009
Reputation:
0
ABSTRACT
The deficiencies of traditional password-based access systems have become more acute as these systems have grown in size and scope. Researchers are actively investigating ways to improve the security of password systems or offer replacements. One category of improvements uses keystroke biometrics, an approach which seeks to identify an individual by their typing characteristics. Since 1980, a number of techniques have been proposed for accurately harnessing keystroke dynamics for system authentication and other novel uses. But do these systems deliver on their promise to increase system security and simultaneously ease the burden of logging into systems and remembering passwords? And do databases of users' keystroke profiles present additional privacy concerns?
The keystroke bio metrics is used with the application in news reporting system. It will detect the person who send the news is the reporter or some other person who hacked the user name and password of the system. First the pattern of the reporter is stored with the server system. Server after receiving the text then it matches with the text pattern information in it.
INTRODUCTION
As an attempt to develop more powerful authentication system, with low cost and good acceptance by users, we proposed here an authentication mechanism based on biometric information of human typing patterns. At the movement we present a complete description of the architecture with the multi pattern verification units and results of first phase implementation of the mechanism based on single pattern verification unit among Multilayer perceptron, Learning-Vector-Quantization, Self-Organizing Neural Networks and Support Vector machine.
Today, all computer based system claims for more sophisticated mechanisms to guarantee the information security. The fast evolution of communication systems provided us a great volume of information anywhere any time. The security question became proprietary. Making these systems reliable and secure is one of the most important challenges of the communication evolution.
Authentication is the way to correctly verify one person who he or she claims to be. Many research works have been developed on the way to correctly identify somebody. Since old times, the humans try to identify each other correctly. The most traditional way to confirm that some body who he or she claims to be is to verify his or her handwritten signature. In computer systems, similar issue should be considered.
All information systems adopt some king of authentication. The most common mechanism is called user name and password. This mechanism consists of basically of an association between a public information (username normally every one knows it) that uniquely identify the user on the system, and a secret word (no body beyond the user should know it) that confirms that the person associated by that user name is who he or she to be. This mechanism presents some drawbacks that make it very week. Some of its drawbacks are: Persons choose easy to break passwords as family names and birthday dates; persons normally write their passwords in places of easy access; one can easily see one password on steal it with no knowledge of disclosure. On the other hand, to develop an additional authentication mechanism that carries some advantages like low cost; high performance and high acceptability are not easy. User name and password are easy to implement (low cost) and are largely acceptable by the users. That s why still most famous authentication mechanism applied nowadays.
There are three main techniques to verify ones identification: some thing a person knows (a code); some thing a person posses(a card); some ting a person has(a characteristic). All these three techniques can be combined on the way to produce more efficient identification system. Naturally if we apply all the three techniques together a more secure authentication mechanism will be produced. However we still have to evaluate the cost and the acceptance issues involved in establishing a more sophisticated authentication mechanism.
The last technique is based on one s biometrics characteristics. A biometrical system is a pattern recognition system that establishes the authenticity either specific physiological characteristics (some particular structural characteristics such as hand size or iris format and color) or behavioral characteristics (some particular behavioral characteristic such as typing speed or writing pressure) inherent to a user. .
One kind of biometrical behavioral characteristic that can be used to provide a particular identification is the dynamics characteristics of someone s typing or the human typing pattern. Many studies shown that this approach is possible and effective.
Combined with traditional authentication system, the user typing information can be of some help to identify users more precisely. Considering behavioral typing information to authenticate users can be very convenient because not extra hardware is necessary. All the behavioral information can be obtained by software systems, what generally implies lower cost than hardware development. Moreover nothing changes in the way the user authenticates himself, what makes it more acceptable. However the great question that lies on the kind of authentication is how precisely can we verify the user typing behavioral characteristics. Many studies, as mentioned before, have shown that this kind of identification is viable but still far from obtaining satisfactory indices of FAR (False Alarm Rate) and IPR (Impostor pass rate) if compared with other biometric techniques like finger printing. We will show that FAR and IPR go on opposite directions It means if we try to reduce one of these indices the other will grow up.
FAR False Acceptance Rate- indicates how frequently the system rejects the valid users. IPR- Imposer Pass Rate Indicates how often systems accept an imposter as an valid one. Those indices are most common metrics to evaluate the performance of biometric systems on their capacity to make mistakes on a matching pattern.
To evaluate typing characteristics two main measures have to be obtained about ones typing:
Inter key time- The time between two consecutive key releases and press.
Press time- The time between one key press event and key release event.
In our solution we will use a combination Inter key time and press time for effective results.
Lets consider two keys x and a are pressed. Let us consider Ta is the press time for key a. Tx is the press time for x. and Tax is the inter key time between Ta release event and tx press time.
This measure can be obtained through typing data collector that read keyboard events and processes this data for feature analysis.
As an attempt to develop good authentication system we proposed here a mechanism that combines two techniques of identification: something that the person knows( i.e., a password) and some thing that a person has (typing characteristics). In this work we are considering static authentication or authentication performed during the access to the system.
Results presented in this work indicate that combining two or more techniques can bring better results in terms of IPR and FAR. In statistical approach FAR and IPR Decrease 41% and 22% respectively in neural network approach 21% and 6% respectively.
The basic idea behind the model proposed here is the support some results presented by Brown and Roogers, which show that some techniques used successfully to verify one s typing pattern could present no good result for other person. It means for some users some techniques would give best results than others.
So an authentication mechanism working with a large group of users to be authenticated should obtain best global results in terms of FAR and IPR if it combines different techniques of typing pattern verification, and uses the same technique for each user.
Our purpose is to implement a mechanism of authentication with 3 techniques and combine it with a cluster analysis to segment users in 3 groups. Each group will be analyzed by one pattern matching technique that should give best results for this group of users.
The authentication mechanism proposed here can operate in two modes: new user registration and user authentication. In the first case, the mechanism will record the user username, password and typing profile.
The typing profile is then analyzed and stored so that it can be used during the authentication phase. In this mode, the user will be asked to type his user name and password about 10 times.
On user verification mode the system will verify if user name and password are correct. After authentication has been done, the typing characteristics of that user are compared to his/her stored profile. If it s typing pattern is quite similar to the stored profile the user is granted access. So the user identifies him self or her self by a user name, a password and the typing characteristics.
The authentication mechanism has five main bocks: registration, password verification, identification, typing verification and decision.
Registration Block: This modules is responsible for registering the user in the system, It begins by obtaining user data {UN- user name, PW password, RS-10 Typing Reference Samples}. After that, a cluster analysis is run to identify the best group for the new user. After the definition of the cluster, each verification unit should be prepared to evaluate that user pattern. Although we are defining a cluster to choose which technique should be used for each user, we will prepare all 3 verification units to analyze any user pattern. Our intention is to keep the operating a multiple verification system and compare its performance with our best verification technique in a future study. For each verification unit, after the training sessions, the registration block will produce a configuration matrix (M) for each verification unit to be able to do the authentication of specific user. In this way, after registration, each user will have on profile database: the username(UN), password(PW) and 3 matrix of configuration (M)- One for each verification unit- and cluster © he or she belongs to.
Password Verification: during the authentication process this block is responsible for searching the username(UN) and password (PW) information on the profile database and for verifying if the username and password combination is correct. If it s not, the user is denied access, other wise the user goes to the next phase of authentication: typing verification.
Identification Block: After a user has been correctly authenticated by password, this block searches for the configuration matrix for this user and the cluster to which him/her each belongs to. So it feeds the next block with the configurations matrix and test typing sample to be verified. If we are working in multiple verification mode, this block will send to next the configuration matrix for each verifier unit. Other wise if we are operating in best verification mode, only the configuration matrix of the verification unit associated to the cluster the user belongs to will be sent.
Verification Block: This block processes the test sample sent by the identification block. To correctly process the sample this block should configure each verifier unit with its specific matrix and so test the sample.
The decision block is responsible for deciding if the result presented by the verification unit makes that test sample valid or not for that user. Generally it will decide based on the threshold that will represent the sensibility of global verifier. If this threshold is high, or make easy to pass through the system, its expected to have low FAR and high IPR. But if not its expected to have high FAR and low IPR. That s big tradeoff of any biometric system; if FAR increases IPR decreases.
If more than one unit is used to process the test sample, the decision block should analyze all results and decide in some way if it is or not a valid sample.
Posts: 58
Threads: 27
Joined: Feb 2009
Reputation:
0
Keystroke authentication is a new access control system to identify legitimate users via their typing behavior. Here, the machine learning techniques have been adapted for the purpose of authentication. The differnt keystroke patterns of the user are created using seven different models and is used to differentiate between the users. Decision Tree, Naive Bayesian, Instance Based Learning, Decision Table, One Rule, Random Tree and K-star etc are the learning methods used for this purpose.
access request, information extraction and authenticating are the three processes involved in the authenticating of a user. The dominant form of authentication nowadays is the Textual based authentication system. But this is a weak system with a high risk of information leak. The Biometric authentication systems is expected to have a better scalability. physiological biometric and behavioral biometrics" are the two different types of biometric technologies used, of which the behavioral biometrics is more flexible. The typical behavioral biometric characteristics include the voice tune and typing pattern.
The authentication using the typing pattern detection is based on the fact that individuals type in a characteristic way on a keyboard.
For more details see:
http://wasetjournals/ijmcs/v1/v1-3-27.pdf
http://bibliotecadigital.sbc.org.br/down...?paper=319
Posts: 50
Threads: 19
Joined: Oct 2009
Reputation:
0
This thread speaks about the clustering techniques used in the biometrics field:
http://seminarsprojects.net/Thread-a-com...biometrics
Decision Trees, a Na ve Bayesian Classifier, Memory Based Learning, and statistics-based Keystroke Dynamics etc are the methods used for this.
This pdf talks about the clustering techniques used in the biometrics field:
http://springerlinkcontent/7v7lja8kkb3rh7xd/
http://wileylegacy/wileychi/eac/pdf/A5204-W.PDF
Posts: 49
Threads: 17
Joined: Oct 2009
Reputation:
0
Thank you sir for sending me mail..i need more information ..
which type of clustering methods are used in user authentication with typing pattern..plz send me mail if u have any inforamation regarding that topic..
Posts: 55
Threads: 23
Joined: Jul 2009
Reputation:
0
plz send me if u know any information about this topic..
plz send me a lot of information about dis topic
Posts: 58
Threads: 21
Joined: Aug 2009
Reputation:
0
hey already i have that information, if u have more information plz mail to me,
[email protected]
|