JP5160911B2 - User authentication device, user authentication method, and user authentication program - Google Patents

Description

  The present invention relates to a personal authentication device, a personal authentication method, and a personal authentication program for estimating the probability of whether or not a user who has accessed by operating a terminal connected to a line is the user.

  2. Description of the Related Art Conventionally, for example, when a user who operates a terminal connected to a line logs in to a membership home page via a Web browser and receives a service, a personal authentication system that accepts an ID and a password from the user and authenticates the user is provided. Although it has been used for a long time, passwords are easy to be fraudulent by third parties, so there is a high risk of impersonation by fraudsters. Against this background, multi-factor authentication that combines a plurality of effective authentication methods in addition to conventional ID and password authentication methods has been proposed.

  Currently, the service provider authenticates the user using a method that uses knowledge that the user knows such as a personal identification number and a password, and a method that uses the characteristics of the person himself such as biometric information and behavioral characteristics, Methods using personal belongings such as IC cards and hardware tokens (tokens storing confidential information) are known, and it is desirable to authenticate by combining these three methods.

  However, although the method using biometric information is powerful as a means for confirming the identity, complicated processing and a dedicated device are required to prevent misidentification and exclusion of others, and this dedicated device is very expensive. In some cases, it is not easy for service providers to introduce. In addition, since the user (end user) needs to set up a dedicated device, the introduction is not easy. Similarly, in a method using a hardware token or the like, it is necessary to distribute a device or the like that reads the token to the user in advance, and the initial introduction cost tends to increase.

  For this reason, as an authentication method that does not require the introduction or distribution of a new device, a method for determining a device using a plurality of attribute information that can be acquired from an existing device (terminal operated by a user) has been proposed. In addition to device attribute information, access information such as connection source network information and access time zone is acquired, and whether or not the user is identified by analyzing and modeling user fragmentary behavior patterns Risk-based authentication methods for determining the risk are also known.

  This risk-based authentication method analyzes the user's behavior pattern from the access log at the time of login, etc., calculates the risk value of each transaction based on it, and determines the authentication method and the risk level based on the risk level based on the policy of the service provider. The authentication strength can be controlled. As a risk-based authentication solution, VeriSign Japan's online fraud detection system is known, and for each factor used for calculating the risk level (for example, IP address, browser information, device information, etc.) A risk value is calculated after defining a policy or rule and setting a threshold (see, for example, Non-Patent Documents 1 and 2).

“VeriSign Identity Protection Online Fraud Detection Service”, [online], [Searched on December 6, 2007], Internet <http://www.verisign.co.jp/vip/resources/pdf/ds_fds2006.pdf> RSA Security, “RSA Adaptive Authentication for Web” (Japanese translation), November 2006 edition, p.1-8

  However, among the conventional techniques described above, the method of performing personal authentication using only device attribute information as a determination element has a risk of device theft and loss, so that impersonation by a third party cannot be completely prevented. There is.

  In other words, when a person other than a regular user requests authentication using a terminal and ID owned by a regular user, or through a network or line used by a regular user and a terminal owned by a regular user And a person other than a regular user who requests authentication (regardless of whether or not a person other than a regular user has malicious intent, for example, a family share a terminal and use an ID In the case of a regular user whose network to be used frequently fluctuates, when a person other than the regular user requests authentication using a terminal and ID owned by the regular user, Spoofing by a third party cannot be completely prevented.

  In addition, among the above-described conventional techniques, in a method of performing personal authentication using a plurality of determination elements (multi-factor authentication method), the dependency between individual determination elements is considered, and the probability of being a regular user is high. There is a problem that it is not possible to determine how much.

  In other words, in the multi-factor authentication method, a decision element that can be quantified by taking a logical product of a plurality of decision elements and finally making an OK or NG decision, or defining a policy for each decision element Needs to be comprehensively determined by setting a threshold. However, in the former determination process, it is difficult to grasp the risk level, such as the probability of being a regular user, and detailed control such as strengthening identity verification according to the service content is not possible. On the other hand, in the latter determination process, the risk level can be expressed, but the determination with high accuracy in consideration of the dependency relationship between the individual determination elements cannot be performed.

  Furthermore, among the above-described conventional techniques, the risk-based authentication solution has a problem that it is not possible to easily realize personal authentication suitable for the service content or to perform high-precision personal authentication.

  In other words, the risk-based authentication solution has the merit that business providers such as service providers can flexibly set rules for calculating risk values, but it is optimized if there is no tuned setting or know-how in the service content This is not easy, and it is not always possible to easily realize identity authentication adapted to the service content. In risk-based authentication solutions, factors for determination (eg, IP address, browser information, device information, etc.) are obtained in pieces, and continuous information representing individual characteristics is acquired. Therefore, for example, a high-risk regular user who is often accessed from a different line or terminal each time may be determined not to be a person, and high-precision personal authentication cannot be performed.

  Therefore, the present invention has been made to solve the above-described problems of the prior art, and can be determined as a quantitative probability value taking into account the dependency relationship between the determination elements, and personal authentication suitable for the service can be easily performed. It is another object of the present invention to provide a personal authentication device, a personal authentication method, and a personal authentication program that can be realized with high accuracy.

In order to solve the above-described problems and achieve the object, the present invention is a personal authentication device that estimates the probability of being a user for a user who has accessed by operating a line-connected terminal, The behavior characteristic information acquisition means for acquiring the behavior characteristic information of the user who operates the terminal from the terminal connected to the line, the behavior characteristic information acquired by the behavior characteristic information acquisition means, and the user who has actually accessed And a probability value calculating means for calculating a probability value indicating the probability that the user is the person using the probability model.

Also, in the present invention according to the above invention, the behavior characteristic information acquisition unit acquires, as behavior characteristic information, custom information indicating customary characteristics related to a user's operation on a terminal operated by each user. It is characterized by doing.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the behavior characteristic information acquisition means acquires terminal information indicating an attribute of a terminal operated by each user as the behavior characteristic information.

Further, the present invention is the above invention, wherein the behavior characteristic information acquisition unit acquires, as behavior characteristic information, line information indicating an attribute of a line to which a terminal operated by each user is connected. Features.

Further, according to the present invention, in the above invention, behavior characteristic information storage means for storing the behavior characteristic information acquired by the behavior characteristic information acquisition means for each user, and behavior characteristic information stored by the behavior characteristic information storage means And a probability model generating means for generating a probability model for each user.

Further, the present invention is the above invention, further comprising authentication means for authenticating whether or not the user who has actually accessed is the user based on the probability value calculated by the probability value calculation means. Features.

Further, according to the present invention, in the above invention, as a result of authenticating whether or not the user who has actually accessed is the user based on the probability value calculated by the probability value calculating means, When the user who has accessed is authenticated as not being the user, the authentication is performed again using another authentication method.

The present invention is also a personal authentication method for estimating a probability of whether or not a user who has been accessed by operating a line-connected terminal is a user, wherein the terminal is connected to the line-connected terminal. Using the behavior characteristic information acquisition step for acquiring the behavior characteristic information of the user to be operated, the behavior characteristic information acquired by the behavior characteristic information acquisition step, and the probability model of the user who has actually accessed the user, And a probability value calculating step of calculating a probability value indicating the probability of identity.

Further, in the present invention according to the above invention, the behavior characteristic information acquisition step acquires, as behavior characteristic information, custom information indicating customary characteristics related to a user's operation on a terminal operated by each user. It is characterized by doing.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the behavior characteristic information acquisition step acquires terminal information indicating an attribute of a terminal operated by each user as the behavior characteristic information.

The present invention, in the above invention, the behavioral characteristic information acquisition step, as the action characteristic information, to acquire the line information indicating the attributes of the line terminals each user is operating is connected Features.

Further, according to the present invention, in the above invention, the behavior characteristic information storage step of storing the behavior characteristic information acquired by the behavior characteristic information acquisition step in a predetermined storage unit for each user, and the behavior characteristic information storage step And a probability model generation step of generating a probability model of each user from the stored behavior characteristic information.

Further, the present invention, in the above invention, further includes an authentication step of authenticating whether or not the user who has actually accessed is the user based on the probability value calculated by the probability value calculation step. Features.

Further, in the present invention according to the above invention, as a result of authenticating whether or not the user who has actually accessed is the user based on the probability value calculated by the probability value calculating step, When the user who has accessed is authenticated as not being the user, the authentication is performed again using another authentication method.

Further, the present invention is a personal authentication program for causing a computer to execute a personal authentication method for estimating a probability of whether or not a user who has accessed by operating a terminal connected to the line is a user. The behavior characteristic information acquisition procedure for acquiring the behavior characteristic information of the user who operates the terminal, the behavior characteristic information acquired by the behavior characteristic information acquisition procedure, and the probability model of the user who has actually accessed And a probability value calculation procedure for calculating a probability value indicating the probability that the user is the user.

Further, in the present invention according to the above invention, the behavior characteristic information acquisition procedure acquires, as behavior characteristic information, custom information indicating customary characteristics related to a user's operation on a terminal operated by each user. It is characterized by doing.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the behavior characteristic information acquisition procedure acquires terminal information indicating an attribute of a terminal operated by each user as the behavior characteristic information.

Also, in the present invention according to the above invention, the behavior characteristic information acquisition procedure acquires, as behavior characteristic information, line information indicating an attribute of a line to which a terminal operated by each user is connected. Features.

Further, according to the present invention, in the above invention, the behavior characteristic information storage procedure for storing the behavior characteristic information acquired by the behavior characteristic information acquisition procedure in a predetermined storage unit for each user, and the behavior characteristic information storage procedure A probability model generation procedure for generating a probability model for each user from the stored behavior characteristic information is further caused to be executed by a computer.

Further, according to the present invention, in the above invention, the computer further executes an authentication procedure for authenticating whether or not the user who has actually accessed is the person himself / herself based on the probability value calculated by the probability value calculation procedure. It is characterized by that.

According to the present invention, in the above invention, the authentication procedure is based on the probability value calculated by the probability value calculation procedure. When authenticating that the user is not the user, the authentication is performed again using another authentication method.

According to the present invention, the probability of whether or not a user who has been accessed by operating a line-connected terminal is the user is estimated, and the behavior characteristic information of the user who operates the terminal from the line-connected terminal Since the probability value indicating the probability that the user is the user is calculated using the acquired behavior characteristic information and the probability model of the user who has actually accessed, it is determined as a quantitative probability value. In addition, it is possible to provide a personal authentication device, a personal authentication method, and a personal authentication program capable of easily and accurately realizing personal authentication suitable for the service.

According to the present invention, as behavior characteristic information, the custom information indicating the custom characteristics related to the user's operation on the terminal operated by each user is acquired, so the custom information indicating the behavior characteristics of the user Therefore, the risk of impersonation when the terminal information is fraudulent, imitated or reproduced can be reduced.

According to the present invention, terminal information indicating the attributes of the terminal operated by each user is acquired as the behavior characteristic information, so whether the same terminal as the terminal at the previous access is used from the information of the user terminal. It can be determined whether or not.

According to the present invention, the line information indicating the attribute of the line to which the terminal operated by each user is connected is acquired as the behavior characteristic information. Therefore, the terminal information is obtained from the line information connected to the user terminal. It is possible to reduce the risk of impersonation when fraudulent, imitated or reproduced.

According to the present invention, the acquired behavior characteristic information is stored for each user, and the probability model of each user is generated from the stored behavior characteristic information, so that it is optimized for individual behavior from past access information It is possible to dynamically create model data for personal identification.

According to the present invention, based on the calculated probability value, it is authenticated whether or not the user who has actually accessed is the person himself / herself. Therefore, the unauthorized access made by another person is detected, and the terminal information is fraudulent and imitated. Or the risk of impersonation when it is reproduced can be reduced.

According to the present invention, based on the calculated probability value, as a result of authenticating whether or not the user who has actually accessed is the principal, if the user who has actually accessed is not the principal, Since the authentication is performed again using the authentication method, even if it is determined that the legitimate access made by the person is unauthorized access based on the calculated probability value, the opportunity for authentication can be given again. Is possible.

  Exemplary embodiments of a personal authentication device, a personal authentication method, and a personal authentication program according to the present invention will be described below in detail with reference to the accompanying drawings.

[1: Configuration of personal authentication device]
A configuration of the personal authentication apparatus according to the embodiment will be described. FIG. 1 is a block diagram illustrating the configuration of the personal authentication device according to the embodiment. The personal authentication device 10 is a device that performs personal authentication for a user who has been accessed by operating the user terminal 40 connected to the line. As shown in the figure, the behavior authentication information extraction unit 30 and the control unit 31a are used. , An authentication processing unit 31 including a prior determination processing unit 31b, a probability model generation unit 31c, and a probability inference unit 31d, and a storage unit 20 including a rule definition unit 21 and a user information unit 22.

  Note that the user terminal 40 is, for example, a personal computer, a home game machine, an Internet TV, an information appliance, a PDA, or a mobile communication terminal such as a mobile phone or a PHS, and at least a communication unit for connecting to a network And an input unit for receiving various operation inputs (for example, a login ID and a login password) from the user.

  Among the units of the personal authentication device 10, the storage unit 20 stores data and programs necessary for various processes by the authentication processing unit 31, and includes a rule definition unit 21 and a user information unit 22, but is closely related to the present invention. As a related item, a user information unit 22 is provided.

  The rule definition part 21 memorize | stores the rule used for the determination process by the prior determination process part 31b mentioned later. To give specific examples, access by the same user ID from a physically distant place in a short time, continuous access for a short time that is impossible for human beings, etc. are clearly judged as unauthorized access. The script description for determining the action to be performed is stored.

  The user information unit 22 stores information on the accessing user. Specifically, as illustrated in FIG. 2, the user information unit 22, in addition to the access history information, from a user terminal 40 connected to the line, an authentication processing unit 31 to be described later performs a user's special operation. Is the behavior characteristic information of the user who operates the user terminal 40 acquired without accompanying the terminal information indicating the attribute of the user terminal 40 operated by each user. Line information indicating the attributes of the connected lines and custom information indicating customary characteristics relating to user operations on the user terminal 40 are stored as access logs.

  As the “access history information”, the user information unit 22 stores, for example, a personal identification ID or login ID, a login password, or the like, which is information input by the user at the user terminal 40.

  Further, the user information unit 22 includes, as “terminal information”, for example, OS information for specifying the OS and browser operating in the user terminal 40, browser type information, browser version information, browser setting information (language setting, etc. ), CPU information specifying the CPU mounted on the user terminal 40, screen resolution specifying the screen of the user terminal 40, system language information set in the user terminal 40, and the like.

  In addition, the user information unit 22 includes, for example, access source IP address information, line identification information uniquely given by the line operator or provider with which the access source line has a contract, Stores provider information (information derived from the IP address information) for specifying the network operator or provider, geographical information of the access source (information that can be derived from the IP address information), and the like.

  The “custom information” includes, for example, time information including an access date, an access time zone, an average keystroke speed of a keyboard when inputting an ID or a password, an adjacent keystroke speed, mouse operation, etc. The trajectory information and the like are stored.

  The behavior characteristic extraction unit 30 acquires, as an access log, behavior characteristic information of a user who operates the terminal from a terminal connected to the line. Specifically, the behavior characteristic extraction unit 30 acquires behavior characteristic information from the user terminal via the network, and stores the acquired behavior characteristic information in the user information storage unit 22. Note that these behavior characteristic information can be acquired from an HTTP header, Java (registered trademark) Script, Flash, ActiveX, or the like incorporated in a Web page, but the acquisition method is not limited thereto.

  The control unit 31a has an internal memory for storing a control program such as an OS (Operating System), a program that defines various processing procedures, and necessary data, and executes various processes using these. Data transmission / reception is performed among the behavior characteristic extraction unit 30, the prior determination processing unit 31b, the probability model generation unit 31c, and the probability inference unit 31d.

  When there is an access from a terminal, the prior determination processing unit 31b may perform a process of determining in advance whether the access is unauthorized access. Specifically, when there is an access from a terminal, the prior determination processing unit 31b determines in advance whether the access from the terminal is an unauthorized access based on the rules stored in the rule definition unit 21. However, when it is determined that the access is illegal, the login process may be stopped. On the other hand, when the prior determination processing unit 31b determines that the access is not an unauthorized access, the prior determination processing unit 31b notifies the control unit 31a accordingly.

  Here, specific examples of determining unauthorized access include access by the same user ID from a physically distant place in a short time, or continuous for a short time that is physically impossible. Actions that are clearly judged as unauthorized access, such as access, are determined as unauthorized access.

  In the authentication processing unit 31, the control unit 31a receives the result of determining whether or not the unauthorized access is performed in the prior determination processing unit 31b based on the behavior characteristic information extracted from the behavior characteristic information extraction unit 30, and when the access is not unauthorized, Then, the probability reasoning unit 31d calculates a person estimation value indicating whether the user who has actually accessed is the person himself or herself based on the probability model already generated by the probability model generation unit 31c. Furthermore, the control unit 31a receives the calculated person estimation value, and determines whether the calculated person estimation value is equal to or greater than a threshold value. As a result, if the control unit 31a determines that the calculated estimated value is equal to or greater than the threshold, the control unit 31a determines that the user is an authorized user and controls the user to log in to the system for receiving the service. The probability model generation unit 31c updates the conditional probability table as the probability model.

  The probabilistic model generation unit 31c is model data for identity determination optimized for individual behavior based on the access information of the user information unit 22 in which the behavior characteristic information extracted from the behavior characteristic information extraction unit 30 is accumulated. Is created dynamically, and the probability model is updated.

  In addition, when the control unit 31a determines that the calculated estimated value is not equal to or greater than the threshold value, additional authentication for performing identity verification is performed again. As a result, when the control unit 31a can confirm the validity, control is performed so that the user can log in to the system for receiving the service, and the probability model is updated. On the other hand, when the validity cannot be confirmed as a result of the additional authentication, the login process is stopped. As an example of additional authentication, there is a method of re-identifying through another communication channel such as a mobile phone or e-mail, but the additional authentication method and threshold criteria can be flexibly set according to the service provider's policy. Can be changed.

  The probability reasoning unit 31d estimates the probability of whether or not the person is the person using the probability reasoning method. As typical probabilistic reasoning methods, multivariate analysis methods such as cluster analysis and discriminant analysis, covariance structure analysis, decision trees, neural networks, and Bayesian networks are known. Is possible. Below, although the Example using a Bayesian network is shown, it is not limited to this.

  Here, processing for generating a conditional probability table as a probability model of the probability reasoning unit 31d will be described with reference to FIGS. As shown in FIG. 3, in the probability reasoning unit 31d, a Bayesian network model is defined as a probability network model in which the dependency relationship of authentication elements is defined. This Bayesian network model can be expressed by a node that represents each authentication element and a link that defines a dependency relationship between the authentication elements.

  Using such a Bayesian network model, the probability reasoning unit 31d reads past behavior characteristic information (access data) that has already been accumulated from the user information unit 22, and exemplifies it in FIG. 4 according to the Bayesian network model. Create a conditional probability table like this: For example, in the example of FIG. 4, when accessed with line identification ID = N1 and OS information = S1, the probability of being a regular user is “0.95” (95%), and line identification ID = N1, When accessed with OS information = S2, the probability of being a regular user is “0.85” (85 percent).

  That is, when the person estimation value calculation process is described using the examples of FIGS. 3 and 4, the probability reasoning unit 31 d transmits the behavior characteristic information Y of the user X (line identification ID = line) to each node configuring the Bayesian network. N1 and OS information = S1) are input, the probability fluctuation is calculated, and the posterior probability value “0.95” of the node X that performs the identity determination is returned. This value is an estimated value representing the identity of the person.

  Here, a calculation example of the principal probability value will be described in detail with reference to FIG. Although briefly described in FIG. 5, “a1” indicates the person, “a2” indicates the other person, and “b” and “c” indicate predetermined behavior characteristic information.

  As shown in the upper part of FIG. 5, the probability reasoning unit 31 d inputs the behavior characteristic information as an access log held by the user information unit 22 to the probability network model, so that the appearance frequency of the observed value of each authentication element Calculate the probability from The probability reasoning unit 31d holds the calculated probability as a conditional probability table. At this time, the conditional probability table is updated based on the dependency relationship of each data, and the conditional probability is updated. Further, the ratio of “a1” and the ratio of “a2” with respect to the entire access log are calculated as target occurrence probabilities.

  Then, as shown in the lower part of FIG. 5, the probability reasoning unit 31 d calculates the personal probability value when there is an access from the user. That is, in the example of FIG. 5, the probability reasoning unit 31 d, when the behavior characteristic information “b1” and “c2” are given as input values, the conditional probability “0. 048 "and the conditional probability" 0.096 "of a2. From these values, the probability of being the person is 33.3 percent (= 0.048 / (0.048 + 0.096)).

  The personal authentication device 10 is a known personal computer, workstation, PDA, or other information processing device, and includes the behavior characteristic extraction unit 30, the control unit 31a, the prior determination processing unit 31b, the probability model generation unit 31c, and the probability. It can also be realized by installing each function of the inference unit 31d.

[2: Processing by personal authentication device]
Next, processing by the personal authentication device will be described with reference to FIG. FIG. 6 is a flowchart showing the flow of the personal authentication process. As shown in the figure, the behavior characteristic extraction unit 30 of the personal authentication device 10 acquires behavior characteristic information from the user terminal via the network when accessed from the terminal (step S101).

  Then, the advance determination processing unit 31b determines in advance whether the access from the terminal is an unauthorized access based on the rules stored in the rule definition unit 21 (step S102), and determines that the access is unauthorized. If so (Yes at Step S102), the login process is stopped (Step S109). On the other hand, when the prior determination processing unit 31b determines that the access is not unauthorized access (No at Step S102), the probability reasoning unit 31d calculates the person estimation value (Step S103). Then, the controller 31a determines whether the calculated estimated value is equal to or greater than a threshold (step S104).

  As a result, when it is determined that the calculated estimated value is equal to or greater than the threshold value (Yes at Step S104), the control unit 31a determines that the user is a regular user (Step S105), and a system for the user to receive a service. Control so that you can log in. Then, the probability model generation unit 31c updates the probability model (that is, the conditional probability table) (Step S106).

  If it is determined that the calculated estimated value is not equal to or greater than the threshold value (No at Step S104), the control unit 31a performs additional authentication for confirming the identity again (Step S107). As a result, when the validity can be confirmed (Yes at Step S108), the control unit 31a performs control so that the user can log in to the system for receiving the service, and updates the probability model (Step S108). S106). On the other hand, when the validity cannot be confirmed as a result of the additional authentication (No at Step S108), the control unit 31a stops the login process (Step S109).

[3: Effect of the embodiment]
As described above, the identity authentication device estimates the probability of whether or not the user who accesses the terminal connected by operating the line-connected terminal and operates the terminal from the terminal connected to the line. User behavior characteristic information is obtained, and the probability value indicating the probability that the user is the user is calculated using the obtained behavior characteristic information and the probability model of the user who has actually accessed. It is possible to provide a personal authentication device, a personal authentication method, and a personal authentication program that can be determined as a probability value and that can easily and accurately realize personal authentication suitable for the service.

  In addition to the user terminal information and the line information connected to the user terminal, the personal authentication device also obtains custom information representing the behavioral characteristics of the user, so that the terminal information can be fraudulent, imitated or reproduced. The spoofing risk can be reduced.

  In addition, since the personal authentication device uses authentication information that can be collected from the user terminal without special user action as the behavior characteristic information, the initial introduction cost and the distribution cost of the new device are suppressed, and the user is excessively charged. High operability can be provided without burden.

  Also, the personal authentication device stores the acquired behavior characteristic information for each user, and generates a probability model for each user from the stored behavior characteristic information, so it is optimal for individual behavior from past access information It is possible to dynamically create a model data for personal identification.

  Also, the identity authentication device can probabilistically express the dependency of each factor by applying probabilistic reasoning technology such as Bayesian network as the identity determination logic for authentication. It is possible to dynamically calculate the risk value without setting.

[4: Other embodiments]
Although the embodiments of the present invention have been described so far, the present invention may be implemented in various different forms other than the embodiments described above. Therefore, another embodiment included in the present invention will be described below.

(1) Example of introduction to actual service For example, in the present invention, as an example of introduction to an actual service, as shown in FIG. 7, an authentication system such as Internet banking provided by a bank as a service providing apparatus or a credit card is used. Application to fraudulent use detection in a payment system is possible.

  In addition, as shown in FIG. 7, the personal authentication device may be connected to the service providing device in a one-to-one relationship to perform the personal authentication processing, or the personal authentication processing may be performed in a one-to-multiple or multiple-to-multiple connection. May be performed. In this case, user authentication processing may be performed by sharing user information regarding common users.

(2) System configuration, etc. In addition, the processing procedures, control procedures, specific names, information including various data and parameters shown in the above documents and drawings may be arbitrarily changed unless otherwise specified. it can.

  Further, each component of each illustrated apparatus is functionally conceptual, and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. Can be integrated and configured. Further, all or any part of each processing function performed in each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware by wired logic.

  The personal authentication method described in the present embodiment can be realized by executing a program prepared in advance on a computer such as a personal computer or a workstation. This program can be distributed via a network such as the Internet. The program can also be executed by being recorded on a computer-readable recording medium such as a hard disk, a flexible disk (FD), a CD-ROM, an MO, and a DVD and being read from the recording medium by the computer.

  As described above, the user authentication apparatus, the user authentication method, and the user authentication program according to the present invention estimate the probability of being a user of a user who has accessed by operating a line-connected terminal. It is useful, and is particularly suitable for determining as a quantitative probability value taking into account the dependency between the determination elements, and for easily and accurately realizing personal authentication suitable for the service.

It is a block diagram which shows the structure of the personal authentication apparatus which concerns on an Example. It is a figure which shows the example of the information memorize | stored in a user information part. It is a figure which shows the example of the Paysian network which is a probability network model. It is a figure which shows the example of the conditional probability table | surface which is a probability model. It is a figure for demonstrating the example of calculation of a principal probability value. It is a flowchart which shows the flow of a personal authentication process. It is a figure for demonstrating the example of application of a personal authentication apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Personal authentication apparatus 20 Memory | storage part 21 Rule definition part 22 User information part 30 Behavior characteristic information extraction part 31 Authentication process part 31a Control part 31b Prejudgment process part 31c Probability model creation part 31d Probability reasoning part 40 User terminal

Claims (21)

A user authentication device that estimates a probability of whether or not a user who has accessed a terminal connected to a line by accessing the line is connected.
Behavior characteristic information acquisition means for acquiring first behavior characteristic information and second behavior characteristic information of a user who operates the terminal from the line-connected terminal;
For the user who has actually accessed , the first behavior characteristic information and the second behavior characteristic information acquired by the behavior characteristic information acquisition means, and the first behavior characteristic information when the user is the user Probability that indicates the probability that the user is the principal using a probability model that combines the conditional probability that occurs and the conditional probability that the second behavior characteristic information occurs when the user is the principal A probability value calculating means for calculating a value;
Based on the probability value calculated by the probability value calculating means, estimating means for estimating whether or not the user who has actually accessed is the person,
Probability model update means for updating the probability model when it is estimated that the user who has actually accessed by the estimation means is the person himself,
A personal authentication device comprising: The behavior characteristic information acquisition means shows a customary characteristic related to a user's operation on a terminal operated by each user as at least one of the first behavior characteristic information and the second behavior characteristic information. The personal identification device according to claim 1, wherein custom information is acquired. The behavior characteristic information acquisition means acquires terminal information indicating an attribute of a terminal operated by each user as at least one of the first behavior characteristic information and the second behavior characteristic information. The personal authentication device according to claim 1 or 2. The behavior characteristic information acquisition means, as at least one of the first behavior characteristic information and the second behavior characteristic information, line information indicating an attribute of a line to which a terminal operated by each user is connected. The personal authentication device according to claim 1, wherein the personal authentication device is acquired. Behavior characteristic information storage means for storing the first behavior characteristic information and the second behavior characteristic information acquired by the behavior characteristic information acquisition means for each user;
Probability model generation means for generating a probability model for each user from the first behavior characteristic information and the second behavior characteristic information stored by the behavior characteristic information storage means;
The personal authentication device according to claim 1, further comprising: 6. The authentication device according to claim 1 , further comprising authentication means for performing additional authentication for confirming the identity again when the estimation means estimates that the user who has actually accessed is not the identity . personal authentication device according to one. The probability model update unit updates the probability model when the user who has actually accessed is authenticated as a result of the additional authentication performed by the authentication unit. 6. The personal authentication device according to 6. A user authentication method for estimating a probability of whether or not a user who has accessed by operating a terminal connected to a line is an identity,
A behavior characteristic information acquisition step of acquiring first behavior characteristic information and second behavior characteristic information of a user who operates the terminal from the terminal connected to the line;
For the user who has actually accessed , the first behavior characteristic information and the second behavior characteristic information acquired by the behavior characteristic information acquisition step, and the first behavior characteristic information when the user is the user Probability that indicates the probability that the user is the principal using a probability model that combines the conditional probability that occurs and the conditional probability that the second behavior characteristic information occurs when the user is the principal A probability value calculating step of calculating a value;
Based on the probability value calculated by the probability value calculation step, an estimation step for estimating whether or not the user who has actually accessed is the user,
A probability model update step of updating the probability model when it is estimated that the user who has actually accessed by the estimation step is the user,
A personal authentication method characterized by including The behavior characteristic information acquisition step shows, as at least one of the first behavior characteristic information and the second behavior characteristic information, a customary characteristic related to a user's operation on a terminal operated by each user. 9. The personal authentication method according to claim 8, wherein the custom information is acquired. In the behavior characteristic information acquisition step, terminal information indicating an attribute of a terminal operated by each user is acquired as at least one of the first behavior characteristic information and the second behavior characteristic information. The personal authentication method according to claim 8 or 9. In the behavior characteristic information acquisition step, as at least one of the first behavior characteristic information and the second behavior characteristic information, line information indicating an attribute of a line to which a terminal operated by each user is connected. The personal authentication method according to claim 8, wherein the authentication method is acquired. A behavior characteristic information storage step of storing the first behavior characteristic information and the second behavior characteristic information acquired by the behavior characteristic information acquisition step in a predetermined storage unit for each user;
A probability model generation step of generating a probability model of each user from the first behavior characteristic information and the second behavior characteristic information stored in the behavior characteristic information storage step;
The personal authentication method according to claim 8, further comprising: 13. The authentication method according to claim 8, further comprising an authentication step of performing additional authentication for confirming the identity again when it is estimated that the user who has actually accessed is not the identity by the estimation step. The personal authentication method described in one . The probability model updating step updates the probability model when authentication is performed again using another authentication method in the authentication step, and as a result, the user who has actually accessed is authenticated. The personal authentication method according to claim 13. A personal authentication program for causing a computer to execute a personal authentication method for estimating a probability of whether or not a user who has been accessed by operating a terminal connected to a line is,
Behavior characteristic information acquisition procedure for acquiring first behavior characteristic information and second behavior characteristic information of a user who operates the terminal from the line-connected terminal;
For the user who has actually accessed , the first behavior characteristic information and the second behavior characteristic information acquired by the behavior characteristic information acquisition procedure, and the first behavior characteristic information when the user is the user Probability that indicates the probability that the user is the principal using a probability model that combines the conditional probability that occurs and the conditional probability that the second behavior characteristic information occurs when the user is the principal A probability value calculation procedure for calculating a value;
Based on the probability value calculated by the probability value calculation procedure, an estimation procedure for estimating whether or not the user who has actually accessed is the user,
Probability model update procedure for updating the probability model when it is estimated that the user who has actually accessed by the estimation procedure is the user,
A personal authentication program characterized by causing a computer to execute. The behavior characteristic information acquisition procedure shows a customary characteristic related to a user's operation on a terminal operated by each user as at least one of the first behavior characteristic information and the second behavior characteristic information. The personal identification program according to claim 15, wherein the custom information is acquired. In the behavior characteristic information acquisition procedure, terminal information indicating an attribute of a terminal operated by each user is acquired as at least one of the first behavior characteristic information and the second behavior characteristic information. The personal authentication program according to claim 15 or 16. In the behavior characteristic information acquisition procedure, as at least one of the first behavior characteristic information and the second behavior characteristic information, line information indicating an attribute of a line to which a terminal operated by each user is connected is used. The identity authentication program according to claim 15, wherein the identity authentication program is acquired. A behavior characteristic information storage procedure for storing the first behavior characteristic information and the second behavior characteristic information acquired by the behavior characteristic information acquisition procedure in a predetermined storage unit for each user;
A probability model generation procedure for generating a probability model of each user from the first behavior characteristic information and the second behavior characteristic information stored by the behavior characteristic information storage procedure;
The personal authentication program according to claim 15, further causing a computer to execute. The computer according to any one of claims 15 to 19, further comprising: causing the computer to execute an authentication procedure for performing additional authentication for verifying the identity again when it is estimated that the user who has actually accessed is not the identity by the estimation procedure. The identity authentication program according to any one of the above. The probability model update procedure updates the probability model when the user who has actually accessed is authenticated as a result of authentication again using another authentication method by the authentication procedure. The personal authentication program according to claim 20, wherein:

Images