A data processing system for user authentication
By generating random, common-sense questions as target questions, the problem of low system security in existing technologies is solved, and the security of user authentication is improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG MEIRI HUDONG NETWORK TECH CO LTD
- Filing Date
- 2023-02-22
- Publication Date
- 2026-06-12
AI Technical Summary
In existing user authentication methods, the questions issued by the system have fixed answers, which are easily cracked by computer programs, resulting in low system security.
Using an initial text list, a pre-set list of question statement templates, and a processor, the system generates intermediate question statements and intermediate word lists by obtaining an initial list of triples. It then randomly issues common-sense questions as target question statements, which are then verified by the user's input. The computer program cannot obtain the answers through calculation.
This improves the security of the user authentication system, ensuring that users can only enter answers themselves and preventing computer programs from cracking the system.
Smart Images
Figure CN116186658B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of information verification, and in particular to a data processing system for user authentication. Background Technology
[0002] Most existing methods for verifying information do so by comparing results. The system issues questions with fixed answers, such as simple calculation formulas, dynamic numerical verification codes, or sequential clicking of system-provided graphics. Users input their answers according to the system's questions. Verification is successful when the user's input matches 100% of the answers stored in the system; otherwise, verification fails.
[0003] However, the above method also has the following technical problems:
[0004] When the system generates a question, it retrieves the question from the database. However, the questions stored in the database are too simple and do not include those with non-fixed answers. When the user enters the answer, since the questions generated by the system have fixed answers that can be calculated by a computer program, the user can obtain the answer through the computer program instead of by entering it themselves, which leads to low system security. Summary of the Invention
[0005] To address the aforementioned technical problems, the technical solution adopted by this invention is as follows:
[0006] A data processing system for user authentication includes: an initial text list A = {A1, ..., A...} i , ..., A m}, a list of preset question statement templates, a processor, and a memory storing computer programs, wherein A i Given the i-th initial text, i = 1...m, where m is the number of initial texts, when the computer program is executed by the processor, the following steps are implemented:
[0007] S100, for A i Process and obtain A i The corresponding initial triplet list B i ={B i1 , ..., B ij , ..., B in(i)}, B ij ={B 1 ij B 2 ij B 3 ij}, B 1 ij For A i The first initial entity in the j-th initial triplet, B2 ij For A i The second initial entity, B, in the j-th initial triplet. 3 ij For B 1 ij and B 2 ij The initial association between them, j = 1...n(i), where n(i) is A i The corresponding initial number of triples.
[0008] S200, according to B i And a list of preset question statement templates, to obtain B i The corresponding intermediate problem statement list C i ={C i1 , ..., C ij , ..., C in(i)}, C ij ={C 1 ij C 2 ij C 3 ij}, C 1 ij ={C 11 ij , ..., C 1r ij , ..., C 1s ij}, C 2 ij ={C 21 ij , ..., C 2r ij , ..., C 2s ij}, C 3 ij ={C 31 ij , ..., C 3r ij , ..., C 3s ij}, C 1r ij For B 1 ij and B 2 ij The r-th intermediate problem statement in the corresponding list of intermediate problem statements, C 2r ij For B 1 ij and B 3 ijThe r-th intermediate problem statement in the corresponding list of intermediate problem statements, C 3r ij For B 2 ij and B 3 ij The r-th intermediate problem statement in the corresponding intermediate problem statement list, where r = 1...s, and s is the number of intermediate problem statements in the intermediate problem statement list.
[0009] S300, according to C i , get C i The corresponding list of intermediate words D i ={D i1 , ..., D ij , ..., D in(i)}, D ij ={D 1 ij D 2 ij D 3 ij}, D 1 ij C 1r ij The corresponding middle word, D 2 ij C 2r ij The corresponding middle word, D 3 ij C 3r ij The corresponding middle word.
[0010] S400, all C i Process the data to obtain the target question statement list U = {U1, ..., U...} corresponding to A. x , ..., U p}, U x Let x be the x-th target question statement corresponding to A, where x = 1...p, and p is the number of target question statements corresponding to A. The target question statement is the statement for all C... 1r ij C 2r ij and C 3r ij Intermediate problem statements after deduplication.
[0011] S500, all D i Process the data to obtain the target word set U corresponding to U. 0 ={U 0 1, ..., U 0 x , ..., U 0p}, U 0 x ={U 0 x1 , ..., U 0 xy , ..., U 0 xq(x)}, U 0 xy For U x The corresponding y-th target word, y = 1...q(x), q(x) is U x The corresponding number of target words, where the target words are D that meet the first preset condition. 1 ij D 2 ij D 3 ij .
[0012] S600, will any U x Send to the target user.
[0013] S700, when receiving feedback from the target user, U x The corresponding verification information is from U x The target user's keywords are obtained from the corresponding verification information, and the target user's keywords are matched with each U. 0 xy The verification information was compared to confirm that it was normal.
[0014] The present invention has at least the following beneficial effects:
[0015] This invention provides a data processing system for user authentication, comprising: an initial text list, a preset question statement template list, a processor, and a memory storing a computer program. When the computer program is executed by the processor, it performs the following steps: obtaining an initial triplet list; obtaining an intermediate question statement list based on the initial triplet list and the preset question statement template list; obtaining an intermediate word list corresponding to the intermediate question statement list; obtaining a target question statement list; obtaining a target word set; obtaining the target user's keywords and comparing the target user's keywords with each target word to determine that the user's verification information is normal information. It is evident that when the system issues questions, it randomly generates questions that are common-sense and have no fixed answers. When the user enters the answer, the computer program cannot calculate the answer to the common-sense question; therefore, the user must manually input the answer for verification, which improves the system's security. Attached Figure Description
[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0017] Figure 1 This is a flowchart illustrating the execution of a computer program in a data processing system for user authentication, as provided in an embodiment of the present invention. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.
[0020] Example 1
[0021] This embodiment provides a data processing system for user authentication, including: an initial text list A = {A1, ..., A...} i , ..., A m}, a list of preset question statement templates, a processor, and a memory storing computer programs, wherein A i Given the i-th initial text, i = 1...m, where m is the number of initial texts, when the computer program is executed by the processor, it performs the following steps, such as... Figure 1 As shown:
[0022] S100, for A i Process and obtain A i The corresponding initial triplet list B i ={B i1 , ..., Bij , ..., B in(i)}, B ij ={B 1 ij B 2 ij B 3 ij}, B 1 ij For A i The first initial entity in the j-th initial triplet, B 2 ij For A i The second initial entity, B, in the j-th initial triplet. 3 ij For B 1 ij and B 2 ij The initial association between them, j = 1...n(i), where n(i) is A i The corresponding initial number of triples; wherein, as those skilled in the art know, any method of obtaining triples from text in the prior art is within the protection scope of this invention, and will not be described in detail here.
[0023] Specifically, the initial text is text obtained from a knowledge database through web crawling, and the initial text has an initial character count not exceeding a preset character count threshold. As those skilled in the art know, any existing knowledge database falls within the protection scope of this invention, and will not be elaborated further here. For example, the knowledge database is the Wikipedia database.
[0024] Preferably, the preset character quantity threshold is set to 100 to avoid the preset character quantity being too large or too small, which would result in low accuracy in obtaining the initial triplet.
[0025] S200, according to B i And the preset question statement template list, obtain B i The corresponding intermediate problem statement list C i ={C i1 , ..., C ij , ..., C in(i)}, C ij ={C 1 ij C 2 ij C 3 ij}, C 1 ij ={C 11 ij , ..., C 1rij , ..., C 1s ij}, C 2 ij ={C 21 ij , ..., C 2r ij , ..., C 2s ij}, C 3 ij ={C 31 ij , ..., C 3r ij , ..., C 3s ij}, C 1r ij For B 1 ij and B 2 ij The r-th intermediate problem statement in the corresponding list of intermediate problem statements, C 2r ij For B 1 ij and B 3 ij The r-th intermediate problem statement in the corresponding list of intermediate problem statements, C 3r ij For B 2 ij and B 3 ij The r-th intermediate problem statement in the corresponding intermediate problem statement list, where r = 1...s, and s is the number of intermediate problem statements in the intermediate problem statement list.
[0026] Specifically, step S200 also includes the following steps:
[0027] S201, Obtain the preset question statement template list E = {E1, E2, E3}, E1 = {E 11 , ..., E 1r , ..., E 1s}, E2={E 21 , ..., E 2r , ..., E 2s}, E3={E 31 , ..., E 3r , ..., E 3s}, E 1r For the r-th preset question statement template in the preset question statement template list corresponding to the first initial entity and the second initial entity, E 2rE is the r-th preset question statement template in the preset question statement template list corresponding to the initial association relationship between the first initial entity and the second initial entity. 3r This is the r-th preset question statement template in the preset question statement template list corresponding to the initial association relationship between the second initial entity and the first and second initial entities.
[0028] Specifically, each of the preset question statement templates includes two slots to be filled and several preset characters. As those skilled in the art know, any method for obtaining preset question statement templates in the prior art is within the protection scope of this invention, and will not be elaborated here.
[0029] Furthermore, the keyword types of the two slots to be filled in each preset question statement template are different, and the keyword type can be any one of the first entity type, the second entity type, and the initial association relationship type.
[0030] S203, B 1 ij Fill in E 1r In the slots to be filled where the keyword type is the first entity type, B 2 ij Fill in E 1r In the slots to be filled where the keyword type is the second entity type, construct C. 1r ij .
[0031] S205, B 1 ij Fill in E 2r In the slots to be filled where the keyword type is the first entity type, B 3 ij Fill in E 2r In the slots to be filled where the keyword type is the initial association type, construct C. 2r ij .
[0032] S207, B 2 ij Fill in E 3r In the slots to be filled where the keyword type is the second entity type, B 3 ij Fill in E 3r In the slots to be filled where the keyword type is the initial association type, construct C. 3r ij .
[0033] The above describes a process where intermediate question statements are obtained from a pre-defined list of question statement templates. These intermediate statements are then processed to obtain the target question statement. When the system issues a question, the target question statement is randomly generated. The target question statement is a common-sense question with no fixed answer. When the user enters the answer to the target question statement, the computer program cannot calculate the answer. Therefore, the user must enter the answer themselves for verification, which improves the system's security.
[0034] S300, according to C i , get C i The corresponding list of intermediate words D i ={D i1 , ..., D ij , ..., D in(i)}, D ij ={D 1 ij D 2 ij D 3 ij}, D 1 ij C 1r ij The corresponding middle word, D 2 ij C 2r ij The corresponding middle word, D 3 ij C 3r ij The corresponding middle word.
[0035] Specifically, D 1 ij The following conditions must be met:
[0036] D 1 ij =B 3 ij .
[0037] Specifically, D 2 ij The following conditions must be met:
[0038] D 2 ij =B 2 ij .
[0039] Specifically, D 3 ij Meets the following conditions:
[0040] D 3 ij =B 1ij .
[0041] S400, all C i Process the data to obtain the target question statement list U = {U1, ..., U...} corresponding to A. x , ..., U p}, U x Let x be the x-th target question statement corresponding to A, where x = 1...p, and p is the number of target question statements corresponding to A. The target question statement is the statement for all C... 1r ij C 2r ij and C 3r ij The intermediate problem statement after deduplication; can be understood as, all C... i C in 1r ij C 2r ij and C 3r ij Inserting into U and performing deduplication is known to those skilled in the art. Any deduplication method in the prior art is within the protection scope of this invention, and will not be described in detail here.
[0042] S500, all D i Process the data to obtain the target word set U corresponding to U. 0 ={U 0 1, ..., U 0 x , ..., U 0 p}, U 0 x ={U 0 x1 , ..., U 0 xy , ..., U 0 xq(x)}, U 0 xy For U x The corresponding y-th target word, y = 1...q(x), q(x) is U x The corresponding number of target words, where the target words are D that meet the first preset condition. 1 ij D 2 ij D 3 ij .
[0043] Specifically, the first preset condition is:
[0044] If there is a target question statement in the target question statement list that is the same as the intermediate question statement corresponding to the intermediate word, then the intermediate word is obtained as the target word corresponding to the target question statement.
[0045] S600, will any U x Send to the target user.
[0046] S700, when receiving feedback from the target user, U x When the corresponding verification information is obtained from the U x The target user's keywords are obtained from the corresponding verification information, and the target user's keywords are matched with each U. 0 xy The verification information was compared to confirm that it was normal.
[0047] Specifically, the S700 procedure also includes the following steps:
[0048] S701, obtain U x The corresponding verification information contains the target user's keyword V. x As those skilled in the art will know, any method for obtaining keywords from information in the prior art falls within the protection scope of this invention, and will not be elaborated further here.
[0049] S703, V x Input into the preset feature extraction model to obtain V x The corresponding first feature vector list V 0 x ={V 0 x1 , ..., V 0 xg , ..., V 0 xh}, V 0 xg For V x The corresponding g-th first feature vector, g = 1...h, where h is the number of first feature vectors. As those skilled in the art know, any model in the prior art that can obtain the feature vectors of words falls within the protection scope of this invention, and will not be elaborated further here.
[0050] S705, U 0 x Input into the preset feature extraction model to obtain U 0 x The corresponding second feature vector list W x ={W x1 , ..., W xy , ..., W xq(x)}, W xy ={W 1xy , ..., W g xy , ..., W h xy}, W g xy For U 0 xy The corresponding g-th second eigenvector.
[0051] S707, according to V 0 x and W x Get V 0 x Corresponding similarity list E x ={E x1 , ..., E xy , ..., E xq(x)}, E xy For V 0 x with U 0 xy The similarity between them, where E xy Meets the following conditions:
[0052]
[0053] S709, when E x E exists in xy Not less than E 0 When the user's verification information is determined to be normal, E 0 This is the preset similarity threshold.
[0054] Specifically, E 0 The value range of E is [0.9-1]. Preferably, E 0 =0.9, to prevent the similarity threshold from being set too high or too low, which could lead to incorrect verification information for identified users.
[0055] As mentioned above, by obtaining the similarity between keywords and target words, it is possible to determine whether the user's verification information is normal, which helps to improve the accuracy of the identified user's verification information and thus improve the security of the system.
[0056] In another specific embodiment, the following steps are included after step S500:
[0057] S10, for U 0 Process and obtain U 0 The corresponding specified word list G = {G1, ..., G...} e , ..., G f}, G e For U 0The corresponding e-th specified word, e = 1...f, f is U 0 The corresponding number of specified words, where the specified words are pairs of U 0 xy The target words after deduplication can be understood as U 0 All U 0 xy Insert into G and perform deduplication.
[0058] S20, Obtain G from U e Corresponding U x Construct the first specified problem statement list H = {H1, ..., H2} corresponding to G. e H f}, H e ={ He1 H ek H et}, H ek For G e The corresponding k-th first specified question statement, k = 1...t, where t is the number of the first specified question statements, can be understood as obtaining the target words corresponding to the target question statement from U and G. e The same target question statement as G e The corresponding first specified question statement.
[0059] S30, Based on H, obtain the second specified question statement list R = {R1, ..., R2} corresponding to G. e , ..., R f}, R e ={R e1 , ..., R ea , ..., R ec}, R ea For G e The corresponding a-th second specified question statement, where a = 1...c, and c is the number of second specified question statements.
[0060] Specifically, step S30 includes the following steps to obtain R. ea :
[0061] S31, from H e Select z first specified question statements from the list, where z is the preset number of first specified question statements to select. Those skilled in the art can set the number of first specified question statements to select according to actual needs.
[0062] S33, input z first specified questions into the preset semantic fusion model, obtain a first intermediate specified question statement, and use it as R. e One of the Rs ea .
[0063] Specifically, the preset semantic fusion model is a pre-trained neural network model, wherein the training samples of the model are multiple question statements, and the output of the model is a question statement.
[0064] As described above, merging multiple questions into one increases the difficulty for users to verify the information. Furthermore, the merged question remains a common-sense question without a fixed answer. When users enter their answers, the computer program cannot calculate the answers to common-sense questions. Therefore, users must manually enter their answers for verification, which improves the system's security.
[0065] S40, process R to obtain the final problem statement list corresponding to A. As those skilled in the art will know, the method for obtaining the final problem statement list corresponding to A is the same as the method for obtaining the target problem statement list corresponding to A, and will not be described again here.
[0066] S50, process H to obtain the final word set corresponding to the final question statement list. As those skilled in the art know, the method for obtaining the final word set corresponding to the final question statement list is the same as the method for obtaining the target word set corresponding to U, and will not be described again here.
[0067] S60 sends any final question statement to the target user.
[0068] S70, when the verification information corresponding to the final question statement fed back by the target user is received, the candidate words of the target user are obtained from the verification information corresponding to the final question statement, and the candidate words of the target user are compared with the final words corresponding to each final question statement list to determine that the user's verification information is normal information. As those skilled in the art know, the method for determining that the user's verification information is normal information refers to the method for determining that the user's verification information is normal information in Embodiment 1, and will not be repeated here.
[0069] As described above, when the system generates questions, the difficulty of the questions is increased by randomization. However, the questions are common-sense questions without fixed answers. When the user enters the answer to the question, the computer program cannot calculate the answer to the common-sense question. Therefore, the user can only enter the answer for verification, which can improve the security of the system.
[0070] This invention provides a data processing system for user authentication, comprising: an initial text list, a preset question statement template list, a processor, and a memory storing a computer program. When the computer program is executed by the processor, it performs the following steps: obtaining an initial triplet list; obtaining an intermediate question statement list based on the initial triplet list and the preset question statement template list; obtaining an intermediate word list corresponding to the intermediate question statement list; obtaining a target question statement list; obtaining a target word set; obtaining the target user's keywords and comparing the target user's keywords with each target word to determine that the user's verification information is normal information. It is evident that when the system issues questions, it randomly generates questions that are common-sense and have no fixed answers. When the user enters the answer, the computer program cannot calculate the answer to the common-sense question; therefore, the user must manually input the answer for verification, which improves the system's security.
[0071] While specific embodiments of the invention have been described in detail by way of example, those skilled in the art should understand that the above examples are for illustrative purposes only and are not intended to limit the scope of the invention. Those skilled in the art should also understand that various modifications can be made to the embodiments without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.
Claims
1. A data processing system for user authentication, characterized in that, The system includes: an initial text list A = {A1, ..., A...} i , ..., A m }, a list of preset question statement templates, a processor, and a memory storing computer programs, wherein A i For the i-th initial text, i=1……m, where m is the number of initial texts, when the computer program is executed by the processor, the following steps are implemented: S100, for A i Process and obtain A i The corresponding initial triplet list B i ={B i1 , ..., B ij , ..., B in（i） }, B ij ={B 1 ij B 2 ij B 3 ij }, B 1 ij For A i The first initial entity in the j-th initial triplet, B 2 ij For A i The second initial entity, B, in the j-th initial triplet. 3 ij For B 1 ij and B 2 ij The initial association between them, j=1……n(i), n(i) is A i The corresponding initial number of triples; S200, according to B i And the preset question statement template list, obtain B i The corresponding intermediate problem statement list C i ={C i1 , ..., C ij , ..., C in（i） }, C ij ={C 1 ij C 2 ij C 3 ij }, C 1 ij ={C 11 ij , ..., C 1r ij , ..., C 1s ij }, C 2 ij ={C 21 ij , ..., C 2r ij , ..., C 2s ij }, C 3 ij ={C 31 ij , ..., C 3r ij , ..., C 3s ij }, C 1r ij For B 1 ij and B 2 ij The r-th intermediate problem statement in the corresponding list of intermediate problem statements, C 2r ij For B 1 ij and B 3 ij The r-th intermediate problem statement in the corresponding list of intermediate problem statements, C 3r ij For B 2 ij and B 3 ij The r-th intermediate problem statement in the corresponding intermediate problem statement list, where r=1……s, and s is the number of intermediate problem statements in the intermediate problem statement list; S300, according to C i , get C i The corresponding list of intermediate words D i ={D i1 , ..., D ij , ..., D in（i） }, D ij ={D 1 ij D 2 ij D 3 ij }, D 1 ij C 1r ij The corresponding middle word, D 2 ij C 2r ij The corresponding middle word, D 3 ij C 3r ij Corresponding middle words; S400, all C i Process the data to obtain the target question statement list U = {U1, ..., U...} corresponding to A. x , ..., U p }, U x Let x be the x-th target question statement corresponding to A, where x = 1...p, and p is the number of target question statements corresponding to A. The target question statement is the statement that applies to all C... 1r ij C 2r ij and C 3r ij The intermediate problem statements after deduplication; S500, all D i Process the data to obtain the target word set U corresponding to U. 0 ={U 0 1, ..., U 0 x , ..., U 0 p }, U 0 x ={U 0 x1 , ..., U 0 xy , ..., U 0 xq（x） }, U 0 xy For U x The corresponding y-th target word, y=1……q(x), q(x) is U x The corresponding number of target words, where the target words are D that meet the first preset condition. 1 ij D 2 ij D 3 ij ; S600, will any U x Send to the target user; S700, when receiving feedback from the target user, U x When the corresponding verification information is obtained from the U x The target user's keywords are obtained from the corresponding verification information, and the target user's keywords are matched with each U. 0 xy The verification information was compared to confirm that it was normal. The process after step S500 also includes: S10, obtain U 0 The corresponding specified word list G = {G1, ..., G...} e , ..., G f }, G e For U 0 The corresponding e-th specified word, e=1……f, f is U 0 The corresponding number of specified words, where the specified words are pairs of U 0 xy The target words after deduplication; S20, Obtain G from U e Corresponding U x Construct the first specified problem statement list H = {H1, ..., H2} corresponding to G. e H f }, H e ={H e1 H ek H et }, H ek For G e The corresponding kth first specified question statement, k=1……t, where t is the number of first specified question statements; S31, from H e Select z first specified question statements from the list, where z is the preset number of first specified question statements to select; S33, input z first specified questions into the preset semantic fusion model, obtain a first intermediate specified question statement, and use it as R. e One of the Rs ea To obtain the second specified problem statement list R={R1, ..., R2} corresponding to G. e , ..., R f }, R e ={R e1 , ..., R ea , ..., R ec }, R ea For G e The corresponding a-th second specified question statement, where a=1……c, and c is the number of second specified question statements; S40, process R to obtain the final question statement list corresponding to A; S50, process H to obtain the final word set corresponding to the final question statement list; The process after step S50 further includes: determining that the user's verification information is normal information based on the final question statement list and its corresponding final word set.
2. The user authentication data processing system according to claim 1, characterized in that, The S200 procedure includes the following steps: S201, Obtain the preset question statement template list E={E1, E2, E3}, E1={E 11 , ..., E 1r , ..., E 1s }, E2={E 21 , ..., E 2r , ..., E 2s }, E3={E 31 , ..., E 3r , ..., E 3s }, E 1r For the r-th preset question statement template in the preset question statement template list corresponding to the first initial entity and the second initial entity, E 2r E is the r-th preset question statement template in the preset question statement template list corresponding to the initial association relationship between the first initial entity and the second initial entity. 3r The r-th preset question statement template in the preset question statement template list corresponding to the initial association relationship between the second initial entity and the first and second initial entities; S203, B 1 ij Fill in E 1r In the slots to be filled where the keyword type is the first entity type, B 2 ij Fill in E 1r In the slots to be filled where the keyword type is the second entity type, construct C. 1r ij ; S205, B 1 ij Fill in E 2r In the slots to be filled where the keyword type is the first entity type, B 3 ij Fill in E 2r In the slots to be filled where the keyword type is the initial association type, construct C. 2r ij ; S207, B 2 ij Fill in E 3r In the slots to be filled where the keyword type is the second entity type, B 3 ij Fill in E 3r In the slots to be filled where the keyword type is the initial association type, construct C. 3r ij .
3. The data processing system for user authentication according to claim 1, characterized in that, The S700 procedure includes the following steps: S701, obtain U x The corresponding verification information contains the target user's keyword V. x ; S703, V x Input into the preset feature extraction model to obtain V x The corresponding first feature vector list V 0 x ={V 0 x1 , ..., V 0 xg , ..., V 0 xh }, V 0 xg For V x The corresponding g-th first eigenvector, g=1……h, where h is the number of first eigenvectors; S705, U 0 x Input into the preset feature extraction model to obtain U 0 x The corresponding second feature vector list W x ={W x1 , ..., W xy , ..., W xq（x） }, W xy ={W 1 xy , ..., W g xy , ..., W h xy }, W g xy For U 0 xy The corresponding g-th second eigenvector; S707, according to V 0 x and W x Get V 0 x Corresponding similarity list E x ={E x1 , ..., E xy , ..., E xq（x） }, E xy For V 0 x with U 0 xy The similarity between them, where E xy Meets the following conditions: ； S709, when E x E exists in xy Not less than E 0 When the user's verification information is determined to be normal, E 0 This is the preset similarity threshold.
4. The user authentication data processing system according to claim 2, characterized in that, Each of the preset question statement templates includes two slots to be filled and several preset characters.
5. The user authentication data processing system according to claim 4, characterized in that, The keyword types for the two slots to be filled in each preset question statement template are different.
6. The user authentication data processing system according to claim 5, characterized in that, The keyword type can be any one of the following: first entity type, second entity type, or initial association type.
7. The user authentication data processing system according to claim 1, characterized in that, D 1 ij Meets the following conditions: D 1 ij =B 3 ij 。 8. The user authentication data processing system according to claim 1, characterized in that, D 2 ij Meets the following conditions: D 2 ij =B 2 ij 。 9. The data processing system for user authentication according to claim 1, characterized in that, D 3 ij Meets the following conditions: D 3 ij =B 1 ij 。 10. The data processing system for user authentication according to claim 1, characterized in that, The first preset condition is: when there is a target question statement in the target question statement list that is the same as the intermediate question statement corresponding to the intermediate word, then the intermediate word is obtained as the target word corresponding to the target question statement.