System and method for adaptively determining an optimal authentication scheme

By receiving user identity and authentication system policies and combining historical information to generate the optimal authentication solution list, the problem of difficulty in balancing security and convenience in the existing technology is solved, and the user-adaptive authentication solution selection is realized, which improves the The security and convenience of the certification system.

CN113383333BActive Publication Date: 2024-11-05SHIFT4 TECHNOLOGY LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN201980076269.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2018-11-19
Filing Date
2019-10-31
Publication Date
2024-11-05
Estimated Expiration
2039-10-31

AI Technical Summary

Technical Problem

The existing authentication system is difficult to adaptively select the optimal authentication scheme based on user attributes, authentication system attributes and previous authentication attempt history, especially when balancing security and convenience, and contradictory requirements are difficult to meet.

Method used

By receiving user identity and authentication system policies, combining previous authentication attempt history information, a weighted score method is used to generate an optimal authentication scheme list, and a selection is presented to the user, allowing users to select preferences, update user success probability and implicit preferences, and use Clustering models and machine learning models predict user preferences and optimize authentication solutions.

Benefits of technology

It achieves a balance between security and convenience, provides the best authentication solution, improves user experience and system security, and adapts to changes in different users and system policies.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113383333B_ABST
    Figure CN113383333B_ABST
Patent Text Reader

Abstract

Method and system for determining an optimal authentication scheme for an authentication system. Embodiments may include: receiving from an authentication system an identity of a user requesting access to the authentication system and a policy of the authentication system; receiving from a storage device data including historical information about previous authentication attempts made by the user; generating a list including optimal selections of one or more authentication schemes, each scheme including one or more authentication factors, based on the historical information and the authentication system policy; and sending the selected list of one or more authentication schemes to the authentication system.
Need to check novelty before this filing date? Find Prior Art

Description

Field of the Invention

[0001] The present invention generally relates to user authentication systems and more particularly to methods and systems for adaptively determining an optimal authentication scheme. Background of the Invention

[0003] Commercially available methods and systems for authentication are generally adapted to receive an identification of a user and authenticate or verify the user's stated identity to grant the user access to corresponding computing resources and / or data. For example, a user may identify himself to an online server (e.g., an online banking server) by name or identification number (ID), and the server may require the user to perform various tests or procedures (e.g., enter a password) to access the requested data.

[0004] Conventional authentication systems typically require a user to perform one or more authentication procedures in a manner that is independent of or separate from various attributes including attributes of the user and the authentication system, as well as a history of previous authentication attempts. SUMMARY OF THE INVENTION

[0006] What is therefore desired is a system and method that can select or recommend an authentication scheme that may include one or more authentication factors to process a user's access request based on attributes of the user (e.g., user profile and user preferences), attributes of the authentication system (e.g., minimum authentication level), and based on a history of previous authentication attempts (e.g., probability of successful first-time authentication).

[0007] Embodiments of the present invention may be adapted to generate or select an authentication scheme that can satisfy multiple different and possibly conflicting requirements through a set of weighted scores. For example, an online banking server may need to perform a number of inconvenient authentication procedures before granting a user access. In a complementary manner, a user may not prefer to go through any authentication process at all.

[0008] An embodiment may suggest or recommend one or more authentication schemes in an optimal manner, in the sense that the selected one or more authentication schemes may optimally satisfy different and possibly conflicting requirements. Related to the above example, an embodiment may (e.g., as part of a predefined policy) apply a first weight to a security level and a second weight to a convenience level. The optimally selected authentication scheme may be one that provides a higher weighted score, thereby mitigating the seemingly conflicting considerations of security and convenience.

[0009] An embodiment of the present invention may include a method for determining an optimal authentication scheme for an authentication system. The method may include:

[0010] receiving from the authentication system the identity of the user who needs to access the authentication system and the policy of the authentication system;

[0011] receiving data from a storage device, the data comprising historical information regarding previous authentication attempts made by a user;

[0012] generating a list of optimal choices of one or more authentication schemes, each scheme including one or more authentication factors, based on the historical information and the authentication system policy; and

[0013] A selection list of one or more authentication schemes is sent to the authentication system.

[0014] Embodiments of the method may include: presenting, by the authentication system, a list of one or more authentication schemes to a user; prompting the user to select at least one authentication scheme; and storing the user's selection as an explicit preference for authentication.

[0015] The authentication system policy may include a required authentication level, at least one available authentication factor, and an associated at least one a priori score.

[0016] The authentication system policy may further include at least one attribute weight selected from a list including a success probability attribute weight, a convenience attribute weight, a risk attribute weight, and a preference attribute weight.

[0017] According to some embodiments, the historical information may include: for each authentication factor, the number of successful previous authentication attempts made by the user; and for each authentication factor, the number of failed previous authentication attempts made by the user. Generating a list of optimally selected authentication schemes may include: for each authentication factor, calculating the probability of successful authentication of the user based on the historical information.

[0018] According to some embodiments, historical information about previous authentication attempts made by a user may include at least one of: the duration of the authentication process; verification cases of successful fraudulent access; verification cases of failed unauthorized access attempts; and explicit selection of authentication factors by the user.

[0019] According to some embodiments, the method may include: receiving data related to personal profile information of a plurality of users; and grouping the plurality of users into a plurality of clusters in a clustering model based on the received personal profile information, wherein each cluster is characterized by a particular cluster parameter value.

[0020] According to some embodiments, the method may include: generating a plurality of authentication schemes S, wherein each authentication scheme (S i ) is a unique combination of available authentication factors; produces a likelihood matrix Y, where each entry Y i,j Indicates that users in cluster j will prefer authentication scheme S iand the probability that the user will succeed in the first authentication attempt; and given the parameters of a particular cluster j, train a machine learning (ML) model to predict Y i,j .

[0021] It can be inferred that the users of cluster j will prefer authentication scheme S according to at least one of the following: i and the likelihood that the user will succeed on the first authentication attempt: the explicit preference score; the probability of success; the convenience score; and the risk score.

[0022] Embodiments of the method may include: receiving profile information of a user; assigning membership in a cluster of a clustering model to the user based on the profile information of the user; predicting a likelihood matrix Y based on cluster parameter values ​​of the membership and based on training of the ML model; determining an optimal entry Y for the user i,j ; Based on the determined optimal entry Y i,j selecting at least one optimal authentication scheme; and generating a list including the optimal at least one selected authentication scheme.

[0023] Embodiments of the method may include obtaining at least one implicit preference of at least one user for an authentication factor based on the user's membership in a cluster.

[0024] Embodiments of the method may include updating, for each authentication factor, a probability of successful authentication of the user based on the user's cluster membership.

[0025] Embodiments of the method may include updating a list with one or more recommended authentication schemes based on at least one of: an updated probability of successful authentication of the user and the obtained at least one implicit preference.

[0026] Embodiments of the present invention may include a system for determining an optimal authentication scheme for an authentication system. The system may include: a non-volatile memory device in which a module of instruction code is stored; and a processor associated with the memory device and configured to execute the module of instruction code. The processor may be configured to, after executing the module of instruction code: receive from the authentication system the identity of a user who needs to access the authentication system and the policy of the authentication system; receive from the storage device data including historical information about previous authentication attempts made by the user; based on the historical information and the authentication system policy, generate a list of optimal selections including one or more authentication schemes, each authentication scheme including one or more authentication factors; and send a selection list of one or more authentication schemes to the authentication system. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The subject matter which is regarded as the invention is particularly pointed out and distinctly claimed in the concluding portion of the specification. However, the organization and method of operation of the invention, together with objects, features, and advantages thereof, may be best understood from the following detailed description when read in conjunction with the accompanying drawings, in which:

[0029] Figure 1 is a block diagram depicting a computing device that may be included in a system for adaptively determining an optimal authentication scheme, according to some embodiments;

[0030] Figure 2 is a block diagram depicting a system for adaptively determining an optimal authentication scheme, according to some embodiments; and

[0031] Figure 3 is a flow chart depicting a method of adaptively determining an optimal authentication scheme, according to some embodiments.

[0032] It should be understood that for simplicity and clarity of illustration, the elements shown in the figures are not necessarily drawn to scale. For example, for clarity, the size of some elements may be exaggerated relative to other elements. In addition, reference numerals may be repeated between the drawings to indicate corresponding or similar elements where deemed appropriate.

[0033] Detailed description of the invention

[0034] In the detailed description below, many specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be appreciated by those skilled in the art that the present invention can be practiced without these specific details. In other cases, well-known methods, processes and parts are not described in detail to avoid blurring the present invention. Some features or elements described about one embodiment may be combined with features or elements described about other embodiments. For the sake of clarity, the discussion of the same or similar features or elements may not be repeated.

[0035] Although embodiments of the present invention are not limited in this regard, discussions using terms such as, for example, "processing," "computing," "calculating," "determining," "establishing," "analyzing," "checking," and the like may refer to operations and / or processes of a computer, computing platform, computing system, or other electronic computing device that processes and / or converts data represented as physical (e.g., electronic) quantities within computer registers and / or memories into physical quantities similarly represented as registers and / or memories of a computer or other information non-transitory storage media that may store instructions for performing the operations and / or processes. Although embodiments of the present invention are not limited in this regard, the terms "plurality" and "a plurality" as used herein may include, for example, "plurality" or "two or more." The terms "plurality" or "aplurality" may be used throughout the specification to describe two or more components, devices, elements, units, parameters, and the like. When used herein, a term set may include one or more items. Unless explicitly stated, the method embodiments described herein are not limited to a particular order or sequence. Additionally, some of the described method embodiments or elements thereof may occur or be performed simultaneously, at the same point in time, or concurrently.

[0036] Embodiments of the system may be associated with an authentication system (e.g., a bank, an online store, etc.) that may require authentication from a user (e.g., a client computer or its user) in order to access (e.g., via an online connection) any computing resource (e.g., a memory module) or data stored therein.

[0037] Embodiments may be configured to adaptively determine one or more optimal (eg, determined by the system to be likely optimal in a given situation) or suggested authentication schemes that may be required by or for a user to pass, to access an authentication system.

[0038] The selection or determination of one or more schemes may be referred to as adaptive, being an improvement to a commercial system implementing an authentication method, in the sense that it may produce different results after changing at least one of the following:

[0039] Attributes or characteristics of the user profile (e.g., the user's place of residence, financial status, history of previous authentication attempts, etc.);

[0040] The user's explicit preferences (e.g., prior selection of an authentication scheme or factor by the user);

[0041] The user's implicit preferences (e.g., preferences made by similar users);

[0042] The authentication system's policy (e.g., minimum required security level, availability of authentication factors, a priori scores for each available authentication factor, etc.);

[0043] A continuously updated score related to the convenience of using each authentication factor;

[0044] A continuously updated score related to the risk associated with using the authentication factor; and

[0045] A continuously updated score related to the probability of first-time success using a particular authentication factor.

[0046] The authentication scheme may be determined to be optimal by satisfying one or more predefined authentication criteria or a combination thereof. For example, the at least one predefined authentication criterion may include:

[0047] At least one security requirement for the authentication system;

[0048] User preferences; and

[0049] The probability that a user will successfully authenticate.

[0050] An embodiment may assign a weight to each certification criterion and select or determine one or more schemes for certification in an optimal manner to obtain a maximum cumulative weighted score according to the assigned weights.

[0051] For example, embodiments may:

[0052] Obtain a score for each certification criterion (e.g., probability of success, risk score, and convenience score);

[0053] Assign or specify corresponding weights for each certification criterion (e.g., success probability weight, risk weight, and convenience weight);

[0054] generating a cumulative weighted sum for each authentication scheme based on the above scores and weights; and

[0055] As described herein, at least one optimal authentication scheme is selected (eg, in the sense of having a maximum cumulative weighted sum).

[0056] Reference now Figure 1 , which is a block diagram depicting a computing device that may be included in an embodiment of a system for adaptively determining a list or set of optimal authentication schemes (e.g., one or more optimal authentication schemes) according to some embodiments.

[0057] The computing device 1 may include a controller 2 (which may be, for example, a central processing unit (CPU) processor, a chip, or any suitable computing or computing device), an operating system 3, a memory 4, an executable code 5, a storage system 6, an input device 7, and an output device 8. The controller 2 (or one or more controllers or processors, possibly across multiple units or devices) may be configured to perform the methods described herein, and / or to perform or act as various modules, units, etc. More than one computing device 1 may be included in a system according to an embodiment of the present invention, and one or more computing devices 1 may act as components of the system.

[0058] Operating system 3 may be or may include any code segment (e.g., a code segment similar to executable code 5 described herein) designed and / or configured to perform tasks involving coordination, scheduling, arbitration, supervision, control, or otherwise managing the operation of computing device 1 (e.g., scheduling the execution of software programs or tasks or enabling software programs or other modules or units to communicate). Operating system 3 may be a commercial operating system. It will be noted that operating system 3 may be an optional component, for example, in some embodiments, the system may include a computing device that does not require or include operating system 3.

[0059] The memory 4 may be or may include, for example, a random access memory (RAM), a read-only memory (ROM), a dynamic RAM (DRAM), a synchronous DRAM (SD RAM), a double data rate (DDR) memory chip, a flash memory, a volatile memory, a non-volatile memory, a cache memory, a buffer, a short-term memory unit, a long-term memory unit, or other suitable memory unit or storage unit. The memory 4 may be or may include a plurality of possibly different memory units. The memory 4 may be a non-transitory readable medium of a computer or a processor, or a non-transitory storage medium of a computer, such as a RAM.

[0060] The executable code 5 may be any executable code, such as an application, program, process, task, or script. The executable code 5 may be executed by the controller 2, possibly under the control of the operating system 3. For example, as further described herein, the executable code 5 may be an application that may adaptively determine an optimal authentication scheme and / or authenticate a user. Although for clarity, in Figure 1 A single executable code 5 is shown in FIG. 1 , but a system according to some embodiments of the present invention may include multiple executable code segments similar to the executable code 5 , which may be loaded into the memory 4 and cause the controller 2 to execute the methods described herein.

[0061] The storage system 6 may be or may include, for example, flash memory as known in the art, memory internal to or embedded in a microcontroller or chip as known in the art, a hard drive, a CD-Recordable (CD-R) drive, a Blu-ray Disc (BD), a Universal Serial Bus (USB) device, or other suitable removable and / or fixed storage unit. Content may be stored in the storage system 6 and may be loaded from the storage system 6 into the memory 4 where it may be processed by the controller 2. In some embodiments, the storage system 6 may be omitted. Figure 1 For example, the memory 4 may be a non-volatile memory having the storage capacity of the storage system 6. Thus, although shown as a separate component, the storage system 6 may be embedded or included in the memory 4.

[0062] Input device 7 can be or can include any suitable input device, component or system, for example, a detachable keyboard or keypad, a mouse, etc. Output device 8 can include one or more (possibly detachable) displays or monitors, speakers and / or any other suitable output devices. As shown in blocks 7 and 8, any applicable input / output (I / O) device can be connected to computing device 1. For example, a wired or wireless network interface card (NIC), a universal serial bus (USB) device or an external hard drive can be included in input device 7 and / or output device 8. It is to be appreciated that any suitable number of input devices 7 and output devices 8 can be operably connected to computing device 1, as shown in blocks 7 and 8.

[0063] Systems according to some embodiments of the present invention may include, for example, but not limited to, multiple central processing units (CPUs) or any other suitable multi-purpose or special-purpose processors or controllers (e.g., a controller similar to controller 2), multiple input units, multiple output units, multiple memory units, and multiple storage units.

[0064] The following table may be used as a reference for the terms used herein.

[0065]

[0066]

[0067] Table 1

[0068] Now the reference is the block diagram Figure 2 , which depicts a system 10 for adaptively determining an optimal authentication scheme according to some embodiments.

[0069] As described herein, system 10 may be or may include at least one computing device (e.g., Figure 1 Element 1) is configured to implement at least one method for adaptively determining an optimal authentication scheme.

[0070] The authentication system 40 may be or may include at least one computing device (e.g., Figure 1 1), and the user 50 may request or require access to its computing resources or data stored therein from the authentication system 40. For example, the authentication system 40 may be an online banking server, and the user 50 may be an online client computer used by a human user requesting access to a bank account.

[0071] The computing device 1 of the authentication system 40 may be configured to execute a set of instruction codes, after which the authentication system 40 may receive an identification (e.g., name, bank account number, etc.) of the user 50, and may authenticate the identity of the user 50 to grant the requested access. For example, the authentication system 40 may execute at least one authentication process, which may include:

[0072] receiving authentication data (e.g., password, biometric data, etc.) from a user;

[0073] comparing the received data to previously stored data (e.g., a previously stored password stored on a database such as element 410); and

[0074] Based on this comparison, the identity of user 50 is verified or authenticated, as is known in the art.

[0075] Access may include the ability to interact with the system, the ability to purchase items from the system, the ability to transfer money to or from the system, or other actions. For example, authentication system 40 may be a server associated with a banking system, and user 50 may request access to data including details of a bank account. Authentication system 40 may include an online merchant, a credit card company, a data storage system, a company that provides data access to employees, or other systems.

[0076] As is known in the art, the authentication system 40 may receive an identity or identification (e.g., a user name, user ID number, bank account number or other account number, etc.), or a "represented" or "provided" identity (who the user claims to be, or an account to which the user claims to have access), and may require authentication of the identity of the user 50 to grant the requested access. As used herein, an identity may include the identification of an account or entity associated with a user. For example, an identity may include a financial account number to which the user has access. The authentication process may include the use of one or more authentication factors. Each factor may involve at least one identity criterion, commonly referred to as an "is," "has," or "knows" identity criterion.

[0077] For example, the identity criterion may be that the user "is" something (e.g., a person with a particular face, voice, fingerprint, etc.), and the authentication factor may be a biometric authentication with at least one biological element (e.g., face, voice, and fingerprint recognition, respectively).

[0078] In another example, the identity criterion may be something that the user "has" (e.g., a particular computing device, such as a cellular phone with a particular International Mobile Equipment Identity (IMEI) identification number, or a client computer with a particular Media Access Control (MAC) address). The authentication factor may be the user's association with a previously authenticated stored identification number (e.g., a previously registered MAC or IMEI).

[0079] In another example, the identity criterion may be something that the user "knows" (e.g., a password, an answer to a predefined question, etc.), and the authentication factor may include verification of the user against such previously stored data.

[0080] In some embodiments, the authentication system 40 may store a score table 412 of supported or available authentication factors (e.g., password authentication, MAC address authentication, and facial recognition) in the database 410. Table 2 below provides an example of a score table 412 according to some embodiments, which may be included in a system for adaptively determining an optimal authentication scheme.

[0081] Identity Standards Authentication Factor Prior score "Know" password Score 1 "Know" Predefined Question 1 Score 2 "Know" Predefined Question 2 Score 3 "have" Computer MAC Score 4 "have" Email Accounts Score 5 "have" Cellular IMEI Score 6 "yes" First Fingerprint Score 7 "yes" Second fingerprint Score 8 "yes" Speech Recognition Score 9 "yes" Facial Recognition Score 10

[0082] Table 2

[0083] As shown in Table 2, the authentication system 40 may assign a priori scores to each supported or available authentication factor in the scoring table 412. For example, the authentication system 40 (e.g., an online broker service) may support all or some of the authentication factors set forth above in Table 2. The authentication system 40 may assign low scores in table 412 to "know" identity criteria (e.g., assuming the premise that passwords are often hacked), medium scores to "have" identity criteria (e.g., assuming the premise that identification numbers are not easily accessible), and high scores to "is" identity criteria (e.g., assuming the premise that biometric data may be difficult to forge).

[0084] The authentication system 40 (e.g., an online banking server) may include a policy 411, which may be specific or unique to the particular authentication system 40 (e.g., as part of a predefined policy determined by the bank). The policy 411 may be stored (e.g., as a data object on the database 410) and may dictate various features of the authentication process of the authentication system 40.

[0085] In some embodiments, policy 411 may include a minimum required authentication level. The minimum required authentication level may in turn specify one or more possible authentication schemes.

[0086] For example, a minimum required authentication level (e.g., a numeric value "N") may be satisfied by an authentication scheme that may include a combination of authentication factors whose cumulative sum of scores equals or exceeds the minimum required authentication level (e.g., >= N).

[0087] In another example, the policy 411 of the authentication system 40 may require a plurality of identity criteria. Thus, the minimum required authentication level (N) may be met by an authentication scheme that includes a combination of authentication factors for different identity criteria, the cumulative sum of the scores of the combination of authentication factors equaling or exceeding the minimum required authentication level (e.g., >= N).

[0088] According to some embodiments, the system 10 may receive from the authentication system 40 at least one element of the policy 411 of the authentication system 40 (e.g., the required authentication level) and the identity of the user who needs to access the authentication system. Thus, as explained herein, the system 10 may generate a list or set 310 of optimally selected authentication schemes based on the authentication system policy 411 and based on historical information. The possible authentication schemes may be stored by the authentication system 40 on the database 410 as a table 413, and / or presented by the authentication system 40 to the user 50 as part of the authentication process.

[0089] Regarding the same example, policy 411 may specify a minimum required authentication level, which in turn may indicate three possible authentication schemes:

[0090] The first scheme may include three "knowing" authentication factors. For example, the user 50 may be required to provide a password, a pet's name, and a relative's name to accumulate scores for the three authentication factors and obtain the required authentication level.

[0091] The second scheme may include two authentication factors, such as a "knowing" factor and a "having" factor. For example, the user 50 may be required to provide a password from a pre-registered device (e.g., with a previously verified MAC address) to accumulate scores for the two authentication factors and obtain the required authentication level.

[0092] A third approach may include a single "yes" authentication factor. For example, the user 50 may be required to present their fingerprint to obtain the required level of authentication. Table 3 below depicts an embodiment of table 413 according to this example.

[0093]

[0094] Table 3 (Example of possible solution table 413).

[0095] According to some embodiments, policy 411 may include at least one of: a minimum required authentication level and at least one available authentication factor associated with a priori score. The term "available" is used herein to refer to an authentication factor (e.g., fingerprint-based authentication) that can be supported by an authentication system 40 (e.g., an authentication system associated with a fingerprint reader).

[0096] System 10 can receive policy 411 or at least one element thereof from authentication system 40 and can generate a list or set of at least one suggestion (e.g., as shown in Table 3) for an authentication scheme having corresponding one or more authentication factors, the combination of which can satisfy the requirements of policy 411 of authentication system 40 (e.g., minimum authentication level requirements, diversity requirements, etc.).

[0097] According to some embodiments, system 10 may store (e.g., on database 60) data including historical information about previous authentication attempts made by user 50. Such historical data may include, for example, the number of successful or failed previous authentication attempts made by user 50 for each authentication factor.

[0098] For example, a first user 50 may attempt to authenticate their identity via a first authentication factor (e.g., voice recognition-based authentication), and may fail (be rejected by the authentication system 40), and a second user 50 may attempt to authenticate themselves via an authentication scheme that includes one or more second authentication factors (e.g., a password and a MAC address), and may succeed. The system 10 may store historical data on the database 60 that includes the success or failure of each such authentication attempt associated with the identity (e.g., username, email account, etc.) of a particular user 50. The historical data may be made available for reading by the system 10 during future authentication attempts by the same particular user 50.

[0099] In some embodiments, user 50 may attempt to authenticate their identity with respect to authentication system 40, and system 10 may calculate the probability of successful authentication of the user (e.g., the conditions under which system 10 correctly authenticates the identity of user 50) for each available authentication scheme and / or factor based on historical information. For example, processor (e.g., Figure 1 Element 2) can read data related to previous voice recognition-based authentication attempts of the user 50 and extrapolate the percentage of first successful authentication attempts for that authentication factor as a corresponding success probability.

[0100] In accordance with some embodiments, the system 10 may be associated with or may include a violation analysis module 90 that is configured (e.g., by a human supervisor and / or by a machine learning (ML) model) to determine successful fraudulent attempts to authenticate a user's identity and / or failed unauthorized authentication attempts relative to the authentication system 40.

[0101] System 10 may obtain (e.g., from violation analysis module 90) and store (e.g., in database 60) data including risk-related information associated with historical authentication attempts. Such risk-related information may include, for example, the number of verified cases of successful fraudulent authentication attempts per user and per authentication factor, and the number of verified cases of failed unauthorized authentication attempts per authentication factor.

[0102] The system 10 may calculate a risk score based on the stored data and assign it to at least one authentication factor and / or authentication scheme. For example, the system 10 may calculate a percentage of successful fraudulent attempts for a particular authentication factor (e.g., password-based authentication) relative to the total number of authentication attempts using that authentication factor. The risk score may be a function of the calculated percentage (e.g., the percentage of authentication attempts made with the particular authentication factor that do not include successful fraudulent attempts and verified failed unauthorized authentications), and may represent the probability that the particular authentication factor can be safely used (e.g., does not succumb to fraudulent authentication attempts).

[0103] In another example, the following Exponential equation to derive risk score:

[0104]

[0105] in:

[0106] TP is the number of True Positive authentications (i.e., the number of successful authentication attempts by real, non-fraudulent users);

[0107] FP is the number of False Positive authentications (e.g., the number of successful but fraudulent authentication attempts); and

[0108] FN is the number of False Negative authentications (eg, the number of failed authentication attempts by a real, non-fraudulent user).

[0109] According to some embodiments, system 10 may store data associated with the convenience of one or more authentication factors (e.g., in database 60). For example, system 10 may store the duration of one or more authentication attempts (e.g., how long it takes to authenticate the identity of user 50 according to a particular authentication factor), derive a convenience score (e.g., an average of the durations of each authentication attempt), and assign or assign a convenience score to at least one authentication factor.

[0110] According to some embodiments, the system 10 may also store (e.g., in the database 60) data associated with the preferences of one or more users 50 for one or more particular authentication factors. The user preferences may be obtained explicitly (e.g., through direct input from the user 50) or implicitly (as further described below). The user preferences may be added as a factor in the convenience score. For example, when a particular user 50 or a group of particular users 50 attempts to authenticate their identity with respect to the authentication system 40, the authentication factor corresponding to the user preference may be assigned a high convenience score.

[0111] According to some embodiments, the system 10 may be configured to optimally select one or more authentication schemes, each of which includes one or more authentication factors, according to a predefined policy 411 associated with the authentication system 40. For example, the authentication system 40 may be an online banking server, and the corresponding bank may specify a policy for selecting an optimal authentication scheme.

[0112] According to some embodiments, policy 411 may include at least one of the following:

[0113] Minimum number of authentication factors;

[0114] At least one required authentication factor type (e.g., authentication by password and authentication by answering predefined questions);

[0115] At least one required identity criterion (e.g., authentication via “knows,” “is,” and “has” criteria);

[0116] The minimum success probability score for each authentication factor; and

[0117] The minimum overall success probability score for the authentication scenario.

[0118] For example, a policy may dictate that an authentication scheme must:

[0119] including at least two authentication factors, each authentication factor being associated with a different type, such as a facial recognition-based authentication factor and a voice recognition-based authentication factor;

[0120] Facial recognition must have a success rate of at least 99.995%;

[0121] Speech recognition must have a success rate of at least 98%; and

[0122] The overall success probability must be at least 1-(10 ∧ -4).

[0123] The policy 411 may also include one or more attribute weights (e.g., success probability attribute weights, convenience attribute weights, preference attribute weights, and risk attribute weights) associated with corresponding attributes of different authentication factors (e.g., success probability, convenience score, explicit and / or implicit user preferences, and risk scores). In some embodiments, as explained herein, the policy 411 may also include preference attribute weights to include the user's preferences in the selection of the authentication factors.

[0124] Regarding the same example of the bank, the bank's policy 411 may specify a first level of importance for security (e.g., by assigning a first attribute weight to a risk score), a second level of importance for success rate (e.g., by assigning a second attribute weight to a probability of success), and a third level of importance for convenience (e.g., by assigning a third attribute weight to a convenience score).

[0125] Based on historical information and policy 411 (eg, required authentication level, a priori score, and one or more attribute weights), system 10 may generate a set or list 310 of optimal choices of one or more authentication schemes, each including one or more authentication factors.

[0126] Selection of a scheme is referred to herein as optimal in the sense that the system 10 can select or recommend an authentication scheme in a manner that optimizes one or more attribute weights corresponding to a specified policy (e.g., provides the highest value for at least one of: probability of success, risk score, and convenience score).

[0127] In some embodiments, the system 10 may select at least one optimal authentication scheme based on a cumulative sum of the scores weighted by the attribute weights.

[0128] Regarding the same example as bank policy 411, the first attribute weight (e.g., assigned to the risk score) can be higher than the third attribute weight (e.g., assigned to the convenience score). This configuration may require performing additional authentication processes (e.g., an authentication scheme that includes a large number of authentication factors) at the expense of the convenience of the user 50. However, as the number of authentication processes increases, the convenience score may decrease accordingly, and may (based on the corresponding weight limit system 10) affect the cumulative sum and prevent the system 10 from further increasing the number of authentication factors. Therefore, policy 411 can configure the system 10 to balance or mitigate between different authentication requirements (e.g., between the user's preferences and the security requirements of the authentication system).

[0129] The authentication system 40 (e.g., an online banking server) can display the contents of the list or set 310 to the user 50 (e.g., via a web browser) as a list of choices, and can prompt the user 50 to select a solution according to his or her preference. As described herein, the authentication system 40 can propagate the user's 50 selection as feedback to the system 10 for further analysis.

[0130] For example, a first solution including a first authentication factor (e.g., a password) and a second solution including a second authentication factor (e.g., voice recognition) may be provided to user 50. In the first scenario, user 50 may have forgotten their password and may have a smartphone that enables them to record their voice and may therefore prefer the voice recognition solution. In the second scenario, user 50 may not have a smartphone and may therefore prefer the password solution. System 10 may use this information to:

[0131] Fine-tune the selection of future solutions for each specific user 50; and

[0132] As explained herein, the selection of future solutions for each user who is similar (eg, having a similar profile) is fine-tuned to be similar to the selection of future solutions by user 50 .

[0133] The system 10 may select at least one authentication scheme including one or more authentication factors based on data related to the historical information and the policy of the authentication system. For example, such data may include at least one of the following:

[0134] The calculated probability of successful authentication relative to the corresponding attribute weight;

[0135] The calculated convenience score relative to the corresponding attribute weight;

[0136] The calculated risk score relative to the corresponding attribute weights;

[0137] Explicit and / or implicit preference scores relative to corresponding preference attribute ratios (as described below); and

[0138] Cumulative prior points given the minimum required certification level.

[0139] In some embodiments, the selection list 310 of one or more optimal authentication schemes may include an authentication scheme based on the data set forth above. For example, the system 10 may select the optimal scheme that provides the maximum cumulative score based on the above data (e.g., the maximum weighted sum of preference scores, risk scores, convenience scores, etc.).

[0140] Additionally or alternatively, list 310 may include multiple optimal authentication schemes selected or ordered according to their scores (e.g., a maximum weighted sum of scores, a second weighted sum, etc.).

[0141] In another embodiment, the list 310 may include at least one probabilistic selection (e.g., random selection) scheme. Such probabilistic selection can be used to enable the system 10 to diversify feedback from the user 50 (e.g., the user's selection of one authentication scheme from the list or set 310), and, as is known in the art, avoid situations in which the system 10's function of selecting the optimal scheme may be directed toward a local solution rather than a global solution.

[0142] System 10 may send a selected list of one or more authentication schemes 310 to authentication system 40. Authentication system 40 may in turn present list 310 of one or more authentication schemes to user 50 and prompt user 50 to explicitly select one or more authentication schemes of list 310 according to his / her preference.

[0143] According to some embodiments, system 10 can be configured to store (e.g., on database 60) user 50's selections as explicit preferences or selections of authentication schemes and / or authentication factors to use this information in future authentication attempts, and to include aspects of the preferences of user 50 and / or similar users in the selection of an optimal authentication scheme.

[0144] For example, the authentication system 40 (e.g., an online banking server) may present (e.g., via a web browser): a first authentication scheme including at least one first authentication factor (e.g., two "knowing" authentication factors, such as authentication by a password and authentication by answering a predefined question); and a second authentication scheme including at least one second authentication factor (e.g., a single "yes" authentication factor, such as facial recognition). The authentication system 40 may prompt the user 50 to select between the presented schemes (e.g., a selection dialog box on a web browser screen as is known in the art), and record or store (e.g., on a database 410) the user's 50 selection. The authentication system 40 may propagate the user's 50 selection to the system 10, which may in turn count and store (e.g., on a database 60) the number of times each authentication factor has been selected to identify the user's 50 explicit preference.

[0145] According to some embodiments, the system 10 may assign explicit preference scores to each authentication factor based on the frequency of selection of each authentication factor. The system 10 may then generate a list 310 (of optimal selections of authentication schemes and / or authentication factors) based on historical information (e.g., accumulated preference scores) and policies 411 (e.g., required authentication levels and preference attribute weights).

[0146] According to some embodiments, as described herein, the system 10 may include a training subsystem 20 configured to continuously obtain data associated with a user 50 and corresponding authentication attempts from one or more authentication systems 40, and adaptively recommend one or more authentication schemes based on the obtained data.

[0147] According to some embodiments, the system 10 may obtain or receive data related to parameters of personal profile information of a plurality of users 50 from one or more authentication systems 40 and store the profile data (e.g., on a database 60). The data may be obtained, for example, during a user 50 accessing an authentication system 40 or from a database associated with the authentication system 40. For example, at least one authentication system 40 may be an online banking server associated with a bank's database, which may store personal information related to profile parameters of the user. Such parameters may include, for example: the user's age, gender, income, residential address, marital status, bank accounts owned, links to other bank accounts, computing devices owned (e.g., smartphones), email accounts owned, etc.

[0148] The training subsystem 20 may include a segmentation module 210 configured to continuously read or receive data of personal profile parameters of multiple users, and segment or group multiple users 50 into multiple clusters in a clustering model according to the received personal profile information. Such clustering may be implemented by any suitable supervised or unsupervised clustering model.

[0149] For example, the segmentation module 210 may be or may include an unsupervised K-means clustering model configured to cluster a plurality of users according to an optimal number of K clusters around a corresponding number of K centroid vectors. As known in the art, each centroid vector of the clustering model may be characterized by a specific clustering parameter value, which may or may not be equivalent to a personal profile parameter.

[0150] The user 50 may require authentication by the authentication system 40 and may provide personal profile parameters for this purpose. The system 10 may then assign membership in one of the clusters to the user, as described herein, and may generate a list 310 (e.g., optimal selection of one or more authentication schemes) based on the membership of the various users in the clusters.

[0151] According to some embodiments, the segmentation module 210 may be configured to obtain or determine at least one implicit preference of at least one user for an authentication factor based on the user's membership in a cluster.

[0152] For example, the segmentation module 210 may receive personal profile parameters (e.g., ownership of a smartphone) related to a user 50 for whom the preference for authentication factors is unknown (e.g., under conditions where this is the user's first authentication attempt). The segmentation module 210 may associate or assign the membership of the user 50 to a cluster of the user based on the user's personal profile parameters (e.g., similar users who also own smartphones). The segmentation module 210 may determine at least one implicit preference associated with the membership cluster (e.g., a user who owns a smartphone may prefer fingerprint recognition for authentication rather than authentication via sending a password), and assign the implicit preference to the new user.

[0153] According to some embodiments, the system 10 may assign an implicit preference score based on the implicit selection of the at least one authentication factor. The system 10 may update the list or set 310 of at least one authentication scheme based on the implicit preference score so as to present at least one authentication factor to the user based on the explicit selections and preferences of similar users (e.g., users of the same cluster).

[0154] According to some embodiments, the segmentation module 210 can update the probability of successful authentication of the user for each authentication factor based on the user's cluster membership. Regarding the same example, the segmentation module 210 can summarize the probability of successful authentication of the user via the fingerprint authentication method based on the average probability of success of other users in the membership cluster. In a similar manner, the system 10 can update the list or set 310 of at least one authentication scheme based on the updated probability of successful authentication of the user.

[0155] The training subsystem 20 may include a solver module 220 configured to use specific authentication factors to predict user preferences and the likelihood of success on a first attempt at an authentication process.

[0156] The solver module 220 may be or may include a supervised machine learning (ML) model 221 configured to receive the clustering data 211 of the segmentation module 210 and stored data related to historical authentication attempts.

[0157] For example, when the segmentation module 210 is implemented as a K-means clustering model, the solver module 220 may receive corresponding K centroid vectors from the segmentation module 210 as the clustering data 211 .

[0158] The data stored regarding historical authentication attempts may include at least one of the following:

[0159] a risk score associated with a particular authentication factor;

[0160] a success probability associated with the authentication factor;

[0161] A convenience score associated with a particular authentication factor; and

[0162] Explicit and / or implicit selection scores of authentication factors and / or schemes.

[0163] The solver module 220 may be adapted to generate a prediction function 221 adapted to predict an optimal authentication factor and / or authentication scheme based on the clustered data and the collected historical authentication data.

[0164] In other words, solver module 220 may deduce a function that relates profile parameters of at least one user cluster to stored historical data related to previous authentication attempts.

[0165] For example, the solver module 220 can generate a plurality of authentication schemes (e.g., a vector “S” of authentication schemes, each authentication scheme comprising a combination or permutation of authentication factors), wherein each authentication scheme (S i) is a unique combination of available authentication factors. (The term unique is used herein to refer to the case where for each index m and index n, where ((m, n) = 0, 1, 2 ...) and m! = n, there exists S m ! =S n ).

[0166] The solver module 220 may generate a likelihood matrix [Y], where each entry Y i,j Indicates that users in cluster j will prefer authentication scheme S i and the likelihood that the user will succeed on the first authentication attempt.

[0167] For example, the user preferred authentication scheme S of cluster j can be inferred according to at least one of the following: i and the likelihood that the user will succeed in the first authentication attempt: an explicit preference score, an implicit preference score (as described elsewhere), a probability of success, a convenience score, and a risk score. For example, each of the aforementioned scores (e.g., the risk score) can be expressed as a score between 0 and 1 (e.g., the probability that authentication is not failed due to fraudulent access), and the likelihood can be calculated as the product of the aforementioned scores.

[0168] As is known in the art, the machine learning model 221 can be trained to identify the cluster j parameters and the authentication scheme S given a particular cluster j parameter and the authentication scheme S i The predicted likelihood value Y i,j .

[0169] For example, the ML 221 may receive at least one of the following as feedback:

[0170] From an administrative user (eg, a user other than user 50 for whom the authentication system performs the authentication process) via a user interface (eg, via Figure 1 Input element 7) includes, for example, the result of the authentication process, changes in the authentication system 40 policy, changes in available authentication factors, etc.;

[0171] input from the authentication system 40 regarding the user 50's selection of at least one authentication factor;

[0172] input from the authentication system 40 regarding the success or failure of the selected authentication factor; and

[0173] at least one input from the violation analysis module 90 regarding the authentication result (e.g., whether the access was fraudulent),

[0174] and training the ML model 221 based on the feedback to predict or infer the entry Y of the likelihood matrix [Y] given the vector clustering parameters associated with a particular cluster of the clustering model 211 i,jA correlation with at least one value of .

[0175] As is known in the art of machine learning, a variety of algorithms may be applied to infer or train the ML model 221 based on the provided historical data to predict such correlations. For example, the solver module 220 may implement a gradient descent algorithm for this purpose.

[0176] According to some embodiments, the system 10 may receive the user's personal profile information from the authentication system 40 (e.g., in real time or near real time during the authentication process of the user 50). The segmentation module 210 may specify or assign membership in one of the clusters of the clustering model 211 to the user based on the user's personal profile information. The solver module may predict at least one entry Y of the likelihood matrix [Y] based on the cluster parameter value of the membership and based on the training of the ML model 221. i,j The value of .

[0177] According to some embodiments, the system 10 may include a decision module configured to receive at least one entry Y of the likelihood matrix [Y] i,j , and (e.g., given the user membership of a particular cluster and stored data on historical authentication attempts) according to at least one entry Y of the likelihood matrix [Y] i,j At least one optimal authentication factor for use by user 50 is determined.

[0178] Reference now Figure 3 , Figure 3 is a flow chart depicting a method for executing a program executed by at least one processor (e.g., Figure 1 Element 2) adaptively determines the optimal authentication scheme (e.g., Figure 2 method of element 40).

[0179] As shown in step S1005, the processor can be configured to receive from the authentication system the identity (e.g., ID number, name, email account, etc.) of a user requesting access to the authentication system (e.g., an online database server) and a policy of the authentication system (e.g., a minimum authentication level, at least one available authentication factor, at least one attribute weight, at least one prior score, etc.).

[0180] As shown in step S1010, the processor may receive data from a storage device including historical information regarding previous authentication attempts made by the user, including, for example, the number of successful authentication attempts, the number of verified fraudulent authentication attempts, the duration of previous authentication attempts, etc.

[0181] As shown in step S1015, the processor may generate a list or set of optimal selections including one or more authentication schemes, each of which includes one or more authentication factors, based on the historical information and the authentication system policy. As described herein, the selection of one or more authentication schemes may take into account at least one of the following: explicit and / or implicit selection or preference of the user for at least one authentication factor, at least one element of the authentication system's policy, and statistical historical data related to previous authentication attempts, including: risks associated with at least one authentication factor, convenience of using at least one authentication scheme (e.g., duration of authentication and number of authentication factors within the authentication scheme), etc.

[0182] As shown in step S1020, the processor may send a selection list of one or more authentication schemes to the authentication system 40 (eg, Figure 2 310), the authentication system 40 may then prompt the user to select an authentication scheme based on their preferences. The processor may store the user's selection for further analysis to iteratively fine-tune the processor's prediction and / or determination of the optimal authentication scheme.

[0183] As is known in the art, commercially available systems for authenticating a user's identity may be independent of or separate from a variety of attributes, including attributes of the user and the authentication system, and a history of previous authentication attempts. Embodiments of the present invention may improve such techniques by providing an authentication method that may be: adapted to accommodate policies that may be dictated by one or more authentication systems and that may dynamically change over time; adapted to accommodate a user's explicit and / or implicit preferences for authentication methods; the most secure (e.g., having the lowest probability of fraudulent access among multiple authentication methods); and the most convenient (e.g., consuming the shortest duration among available authentication methods).

[0184] Although certain features of the invention have been shown and described herein, many modifications, substitutions, changes and equivalents will now occur to those skilled in the art. It is therefore to be understood that the appended claims are intended to cover all such modifications and changes that fall within the true spirit of the invention.

Claims

1. A method for determining an optimal authentication scheme by at least one processor, the method comprising: receiving from an authentication system an identity of a user requiring access to the authentication system and a policy of the authentication system; receiving data from a storage device, the data comprising historical information related to previous authentication attempts made by the user; generating a list of optimal selections of one or more authentication schemes, each authentication scheme including one or more authentication factors, based on the historical information and a policy of the authentication system; as well as sending a selection list of one or more authentication schemes to the authentication system; Wherein, generating the list further comprises: receiving data related to personal profile information of a plurality of users; grouping the plurality of users into a plurality of clusters in a clustering model based on the received personal profile information, wherein each cluster is characterized by a particular clustering parameter value; Generate multiple authentication schemes S, where each authentication scheme S i is a unique combination of available authentication factors; Generate a likelihood matrix Y, where each entry Y i,j Indicates that users in cluster j will prefer authentication scheme S i and the likelihood that the user will succeed in a first authentication attempt; and Train a machine learning model to predict Y given a particular cluster j parameters i,j .

2. The method according to claim 1, further comprising: presenting, by the authentication system, the list of one or more authentication schemes to the user; prompting the user to select at least one authentication scheme; as well as The user's selection is stored as an explicit preference for authentication.

3. The method according to any one of claims 1 and 2, wherein: The policy of the authentication system includes a required authentication level, at least one available authentication factor, and an associated at least one a priori score.

4. The method according to any one of claims 1 to 3, wherein: The policy of the authentication system further includes at least one attribute weight selected from a list including a success probability attribute weight, a convenience attribute weight, a risk attribute weight, and a preference attribute weight.

5. The method according to any one of claims 1 to 4, wherein: The historical information includes: the number of successful previous authentication attempts made by the user for each authentication factor; and the number of failed previous authentication attempts made by the user for each authentication factor; and wherein the list of authentication schemes for generating optimal selections includes: for each authentication factor, calculating the probability of successful authentication of the user based on the historical information.

6. The method according to any one of claims 1 to 5, wherein: The historical information about previous authentication attempts made by the user also includes at least one of the following: Duration of the certification process; Verification cases of successful fraudulent access; Verified cases of failed unauthorized access attempts; and The user's explicit selection of an authentication factor.

7. The method according to any one of claims 1 to 6, wherein: It is inferred that users of cluster j will prefer authentication scheme S according to at least one of the following: i And the probability that said user will succeed on the first authentication attempt: explicit preference scores; Probability of success; Convenience score; and Risk score.

8. The method according to any one of claims 1 to 7, further comprising: receiving a user's personal profile information; assigning membership in a cluster of the cluster model to the user based on the user's personal profile information; predicting a likelihood matrix Y based on the clustering parameter values ​​of the membership and based on the training of the ML model; Determine the optimal entry Y for the user i,j ; According to the determined optimal entry Y i,j selecting at least one optimal authentication scheme; and A list is generated including at least one selected authentication scheme that is optimal.

9. The method according to any one of claims 1 to 8, comprising: At least one implicit preference of at least one user for an authentication factor is obtained based on the user's membership in a cluster.

10. The method according to any one of claims 1 to 9, comprising: The probability of successful authentication for a user is updated for each authentication factor according to the user's cluster membership.

11. The method according to any one of claims 1 to 10, comprising: The list of one or more recommended authentication schemes is updated based on at least one of: an updated probability of successful authentication of the user and the obtained at least one implicit preference.

12. A system for determining an optimal authentication scheme for an authentication system, the system comprising: a non-transitory memory device in which a module of instruction code is stored; and a processor associated with the memory device and configured to execute the module of the instruction code, after executing the module of the instruction code, the processor is further configured to: receiving from the authentication system an identity of a user requiring access to the authentication system and a policy of the authentication system; receiving data from a storage device, the data comprising historical information related to previous authentication attempts made by the user; generating a list of optimal selections of one or more authentication schemes, each authentication scheme including one or more authentication factors, based on the historical information and a policy of the authentication system; as well as sending a selection list of one or more authentication schemes to the authentication system; Wherein, generating the list further comprises: receiving data related to personal profile information of a plurality of users; grouping the plurality of users into a plurality of clusters in a clustering model based on the received personal profile information, wherein each cluster is characterized by a particular clustering parameter value; Generate multiple authentication schemes S, where each authentication scheme S i is a unique combination of available authentication factors; Generate a likelihood matrix Y, where each entry Y i,j Indicates that users in cluster j will prefer authentication scheme S i and the likelihood that the user will succeed in a first authentication attempt; and Train a machine learning model to predict Y given a particular cluster j parameters i,j .

13. The system according to claim 12, wherein: The processor is configured to: presenting, by the authentication system, the list of one or more authentication schemes to the user; prompting the user to select at least one authentication scheme; as well as The user's selection is stored as an explicit preference for authentication.

14. The system according to any one of claims 12 and 13, wherein: The policy of the authentication system includes a required authentication level, at least one available authentication factor, and an associated at least one a priori score.

15. The system according to any one of claims 12 to 14, wherein: The policy of the authentication system further includes at least one attribute weight selected from a list including a success probability attribute weight, a convenience attribute weight, a risk attribute weight, and a preference attribute weight.

16. The system according to any one of claims 12 to 15, wherein: The historical information includes: the number of successful previous authentication attempts made by the user for each authentication factor; and the number of failed previous authentication attempts made by the user for each authentication factor; and wherein the list of authentication schemes for generating optimal selections includes: for each authentication factor, calculating the probability of successful authentication of the user based on the historical information.

17. The system according to any one of claims 12 to 16, wherein: The historical information about previous authentication attempts made by the user also includes at least one of the following: Duration of the certification process; Verification cases of successful fraudulent access; Verified cases of failed unauthorized access attempts; and The user's explicit selection of an authentication factor.

18. The system according to any one of claims 12 to 17, wherein: The processor is configured to infer that the user of cluster j will prefer the authentication scheme S according to at least one of the following: i And the probability that said user will succeed on the first authentication attempt: explicit preference scores; Probability of success; Convenience score; and Risk score.

19. The system according to any one of claims 12 to 18, wherein: The processor is configured to: receiving a user's personal profile information; assigning membership in a cluster of the cluster model to the user based on the user's personal profile information; predicting a likelihood matrix Y based on the clustering parameter values ​​of the membership and based on the training of the ML model; Determine the optimal entry Y for the user i,j ; According to the determined optimal entry Y i,j selecting at least one optimal authentication scheme; and A list is generated including at least one selected authentication scheme that is optimal.

20. The system according to any one of claims 12 to 19, wherein: The processor is configured to obtain at least one implicit preference of at least one user for an authentication factor based on membership of the user in a cluster.

21. The system according to any one of claims 12 to 20, wherein: The processor is configured to update a probability of successful authentication of the user for each authentication factor based on the user's cluster membership.

22. The system according to any one of claims 12 to 21, wherein: The processor is configured to update a list of one or more recommended authentication schemes based on at least one of: an updated probability of successful authentication of the user and the obtained at least one implicit preference.

23. A non-transitory computer-readable storage medium storing instructions which, when executed by a processor, cause the processor to perform the method according to any one of claims 1-11.