Secure aggregate value calculation system, device, method, and program

The secret aggregate value calculation system generates a common secret flag for each record to securely compute aggregate values, addressing the challenge of maintaining secrecy in grouping and attribute usage information, thus enhancing the security and accuracy of calculations.

WO2024241395A1PCT designated stage expired Publication Date: 2024-11-28NT T INC
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
PCT/JP2023/018816
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-19
Publication Date
2024-11-28

AI Technical Summary

Technical Problem

Existing methods for calculating aggregate values in secure multi-party computation fail to keep both grouping information and the information about which attribute values to use secret, especially when using the same random k attribute values for each group.

Method used

A secret aggregate value calculation system comprising multiple devices that generate a common secret flag for each record, allowing for secure calculation of aggregate values while maintaining the secrecy of grouping information and attribute usage, using flag generation and aggregate value calculation units to perform operations like addition, subtraction, and multiplication without revealing secret values.

Benefits of technology

Enables the calculation of aggregate values while keeping both grouping information and attribute usage information secret, ensuring secure and accurate computations across multiple records.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure JP2023018816_28112024_PF_FP_ABST
    Figure JP2023018816_28112024_PF_FP_ABST
Patent Text Reader

Abstract

This secure aggregate value calculation system comprises a plurality of secure aggregate value calculation devices 1a, ..., 1A, wherein: n records that are input data belong to one of a plurality of groups; each of the records includes secret values of m attribute values respectively corresponding to m attributes; [f] is a secret value of grouping information that is information with regard to the groups to which the records belong, and is a flag that is information representing which k attribute values, of the secret values of the m attribute values included in the records, of which the secret values shall be used to calculate an aggregate value; and there are provided flag generation units 11a, ..., 11A for using [f] to generate a secret value of a flag in common to the records included in each of the plurality of groups, and aggregate value calculation units 12a, ..., 12A for using the generated flags and the input data to calculate aggregate values corresponding to the records.
Need to check novelty before this filing date? Find Prior Art

Description

Secret aggregate value calculation system, device, method and program The disclosed technology relates to cryptographic application technology, and in particular to a technology for calculating aggregate values ​​for each record of secretly grouped data without revealing the input or output, while concealing information about random attributes used for common calculations within the group. There is a method called secure computation as a method for obtaining a specific computation result without restoring an encrypted numerical value. In secure computation, for example, encryption is performed by distributing numerical fragments (secret values) among three secret aggregate value computation devices, and the three secret aggregate value computation devices perform cooperative computation, so that the results of addition / subtraction, constant addition, multiplication, constant multiplication, logical operations (negation, logical product, logical sum, exclusive logical sum), and data format conversion (integer, binary number) can be held in a distributed state, i.e., encrypted state, among the three secret aggregate value computation devices, without restoring the secret value. When calculating an aggregate value for each record of data consisting of n records that have been encrypted using secret computation and further secretly grouped, a method is known in which the calculation is performed using a concealed flag that indicates the boundary between the groups (for example, see Non-Patent Document 1). Koki Hamada, Dai Ikarashi, Ryo Kikuchi, and Koji Chida, “Efficient decision tree training with new data structure for secure multi-party computation”, Proc. Priv. Enhancing Technol., Vol. 2023, No. 1, pp. 343-364, 2023. However, in the technology of Non-Patent Document 1, when attempting to calculate an aggregate value using only the same random k attribute values ​​for each group from the data contained in each record (which is assumed to consist of m attribute values), it was not possible to perform the calculation while keeping both the grouping information and the information about which attribute values ​​to use secret. The disclosed technology has an object to provide a secret aggregate value calculation system, device, method, and program that can calculate an aggregate value while keeping both grouping information and information about which attribute values ​​to use secret. A secret aggregate value calculation system according to one aspect of the disclosed technology is a secret aggregate value calculation system including a plurality of secret aggregate value calculation devices, in which n records that are input data belong to one of a plurality of groups, each record includes secret values ​​of m attribute values ​​corresponding respectively to m attributes, and [f] is a secret value of grouping information that is information about the group to which each record belongs, and [f] is a flag that is information indicating which k attribute value secret values ​​among the m attribute values ​​included in the record are used to calculate the aggregate value, and the system is equipped with a flag generation unit that generates a secret value of a flag common to each record included in each of the plurality of groups using [f], and an aggregate value calculation unit that calculates an aggregate value corresponding to each record using the generated flag and the input data. According to the disclosed technology, it is possible to calculate an aggregate value while keeping both the grouping information and the information about which attribute value to use secret. Fig. 1 is a diagram showing an example of the functional configuration of a secret aggregate value calculation system. Fig. 2 is a diagram showing an example of the functional configuration of a secret aggregate value calculation device. Fig. 3 is a diagram showing an example of a processing procedure of a secret aggregate value calculation method. Fig. 4 is a diagram showing an example of the functional configuration of a computer. Hereinafter, embodiments of the disclosed technology will be described with reference to the drawings. In the drawings, components having the same functions are denoted by the same reference numerals, and duplicated descriptions will be omitted. An example of the configuration of a secret aggregate value calculation system according to an embodiment will be described with reference to FIG. The secret aggregate value calculation system has A (≧2) secret aggregate value calculation devices 11, ..., 1 A In this embodiment, the secret aggregate value calculation devices 1, ..., 1 Aare each connected to a communication network 2. The communication network 2 is a circuit-switched or packet-switched communication network configured so that each connected device can communicate with each other, such as the Internet, a LAN (Local Area Network), or a WAN (Wide Area Network). Note that each device does not necessarily need to be able to communicate online via the communication network 2. For example, the secret aggregate value calculation devices 1, ..., 1 A The information to be input to the secret aggregate value calculation device 1 is stored on a portable recording medium such as a magnetic tape or a USB memory, and the information is read from the portable recording medium to the secret aggregate value calculation device 1. A It may also be configured so that the data can be input offline. Referring to FIG. 2, a secret aggregate value calculation device 1 included in the secret aggregate value calculation system a A configuration example of (a=1, ..., A) will be described. a As shown in FIG. 2, the flag generation unit 11 a and the aggregate value calculation unit 12 a The flag generating unit 11 includes, for example, a The temporary flag generating unit 111 a , selection unit 112 a , replication section 113 a For example, the present invention is provided with: Secret aggregate value calculation device 1 a Each component of (1≦a≦A) is another secret aggregate value calculation device 1 a' The secret aggregate value calculation method of the embodiment is realized by carrying out the processing of each step described below in cooperation with each component (a'=1, . . . , A, where a ≠ a'). The processing of each step is performed by secure computation. a performs the processing at each step without restoring the secret value, in other words, without knowing the contents of the secret value. Secret aggregate value calculation device 1 ais a special device configured by loading a special program into a publicly known or dedicated computer having a central processing unit (CPU), a main memory (RAM), etc. a For example, each process is executed under the control of a central processing unit. a The data input to the secret aggregate value calculation device 1 and the data obtained in each process are stored in, for example, a main memory device, and the data stored in the main memory device is read out to the central processing unit as necessary and used for other processes. a At least a part of each of the components may be configured by hardware such as an integrated circuit. In the following description, α is an arbitrary value or vector, and [α] is a secret value of α. [α] is, for example, a share obtained by secretly sharing α. The processing procedure of the secret aggregate value calculation method executed by the secret aggregate value calculation system of the embodiment will be described with reference to FIG. 2 and FIG. Secret aggregate value calculation device 11,...,1 A Input data is input to . The input data is n records. Each record contains m secret values ​​of attribute values ​​corresponding to m attributes. n is an integer of 2 or more. m is an integer of 2 or more. For example, the i-th (1≦i≦n) record [a[i]]=([a[i,1]],...,[a[i,m]]) consists of m secret attribute values. Each of n records belongs to one of multiple groups. For example, in the input data, records that belong to the same group are arranged consecutively. Let [f] be the secret value of grouping information f, which is information about the group to which each record belongs. For example, [f] = ([f[1]], ..., [f[n]]) is composed of n secret values. For example, if the i-th record of the input data is the first element in the group, then the i-th element of f, [f[i]], has a secret value of 1, [1]; otherwise, the i-th element of f, [f[i]], has a secret value of 0. For example, if n=6 and the 1st and 2nd are in the same group as the 3rd, 4th, 5th, and 6th, then the secret value [f] of the grouping information f is [f]=([1],[0],[1],[0],[0],[0]). Hereinafter, this example will be called Example E. <Step S1> Flag generation unit 11 a ,…,11 A [f] is entered. Flag generation unit 11 a ,…,11 A is a flag that is information indicating which k confidential values ​​of attribute values ​​among the m confidential values ​​of attribute values ​​contained in a record are to be used to calculate an aggregate value, and a confidential value of a flag common to each record contained in each of a plurality of groups is generated using [f] (step S1). The generated confidential value of the flag is calculated by the aggregate value calculation unit 12. a ,…,12 A will be output. For example, the flag generation unit 11 a ,…,11 A are the temporary flag generating units 111. a ,…,111 A and the selection unit 112 a ,…,112 A and the duplication unit 113 a ,…,113 A It is equipped with: In this case, the temporary flag generation unit 111 a ,…,111 A generates a temporary flag secret value for each record, which is information indicating which k attribute value secret values ​​among the m attribute value secret values ​​are used to calculate the aggregate value. The temporary flag secret value may be generated randomly. For example, the temporary flag generation unit 111 a ,…,111 AFor each i-th record, a vector of k secret values ​​of 1 [1] and m?k secret values ​​of 0 [0] is rearranged in a random order using secret computation with random permutation to generate a sequence of secret values ​​[b[i]] = ([b[i,1]], ..., [b[i,m]]), which is used as the secret value of the temporary flag. k is an integer between 1 and m. In example E, when m=5 and k=2, the following secret value [b[i]] of the temporary flag is generated, for example. [b[1]]=([0],[1],[1],[0],[0]), [b[2]]=([0],[0],[1],[0],[1]), [b[3]]=([1],[0],[1],[0],[0]), [b[4]]=([0],[1],[1],[0],[0]), [b[5]]=([0],[0],[1],[0],[1]), [b[6]]=([0],[1],[0],[0],[1]) In addition, the selection unit 112 a ,…,112 A selects one temporary flag secret value from the temporary flag secret values ​​of the records included in each of the multiple groups. For example, the selection unit 112 a ,…,112 A selects the secret value of the temporary flag corresponding to the record corresponding to the element [1] among the elements of [f]. In other words, the selection unit 112 a ,…,112 A selects the confidentiality value of the tentative flag corresponding to the first record in each group. In the above example, since [f]=([1],[0],[1],[0],[0],[0]), the secret value of the temporary flag corresponding to the record corresponding to the first element of [f], [b[1]]=([0],[1],[1],[0],[0]), and the secret value of the temporary flag corresponding to the record corresponding to the third element of [f], [b[3]]=([1],[0],[1],[0],[0]), are selected. Then, the replicating unit 113 a ,…,113 AThe hidden value of the selected temporary flag is set as the hidden value of the flag of each record included in the group that includes the record corresponding to the temporary flag. In the above example, the tentative flag secret value [b[1]] = ([0], [1], [1], [0], [0]) is set as the secret value of the flag of the first and second records. In addition, the tentative flag secret value [b[3]] = ([1], [0], [1], [0], [0]) is set as the secret value of the flag of the third to sixth records. As a result, the flag secret value [d[i]] of the i-th record is as follows: [d[1]]=([0],[1],[1],[0],[0]), [d[2]]=([0],[1],[1],[0],[0]), [d[3]]=([1],[0],[1],[0],[0]), [d[4]]=([1],[0],[1],[0],[0]), [d[5]]=([1],[0],[1],[0],[0]), [d[6]]=([1],[0],[1],[0],[0])) In addition, the duplication unit 113 a ,…,113 A In order to perform the above processing, the concealment values ​​of the provisional flags of the records other than the record corresponding to the concealment value of the selected one provisional flag may be set to a concealment value of 0, and a flag having the sum of the i-th concealment values ​​of the provisional flags within the group as the i-th concealment value may be set as a flag common to each record included in the group. For example, the replicator 113 a ,…,113 A First, perform a secret computation to multiply each element [b[i,1]],...,[b[i,m]] of the secret value [b[i]] of the temporary flag of the i-th record by the i-th element [f[i]] of [f], i.e., perform the computation [c[i,j]]=[b[i,j]][f[i]] to obtain the secret value of the temporary flag [c[i]]=([c[i,1]],...,[c[i,m]]). As a result, in the above example, the following secret value [c[i]] of the temporary flag is generated: [c[1]]=([0],[1],[1],[0],[0]), [c[2]]=([0],[0],[0],[0],[0]), [c[3]]=([1],[0],[1],[0],[0]), [c[4]]=([0],[0],[0],[0],[0]), [c[5]]=([0],[0],[0],[0],[0]), [c[6]]=([0],[0],[0],[0],[0])). Then, the replicating unit 113 a ,…,113 A A flag whose i-th secret value is the sum of the i-th secret values ​​of the temporary flags in the group is set as a common flag for each record included in the group. In the above example, the confidential flag values ​​[d[1]], [d[2]] of each record included in the group consisting of the first and second records are as follows: ([c[1,1]]+[c[2,1]],[c[1,2]]+[c[2,2],[c[1,3]]+[c[2,3]],[c[1,4]]+[c[2,4]],[c[1,5]]+[c[2,5]]) =([0]+[0],[1]+[0],[1]+[0],[0]+[0],[0]+[0]) =([0],[1],[1],[0],[0]) In the above example, the confidentiality values ​​[d[3]], [d[4]], [d[5]], and [d[6]] of the flags of the records in the group consisting of the third to sixth records are as follows: ([c[3,1]]+[c[4,1]]+[c[5,1]]+[c[6,1]],[c[3,2]]+[c[4,2]]+[c[5,2]]+[c[6,2]],[c[3,3]]+[c[4,3]] +[c[5,3]]+[c[6,3]],[c[3,4]]+[c[4,4]]+[c[5,4]]+[c[6,4]],[c[3,5]]+[c[4,5]]+[c[5,5]]+[c[6,5]]) =([1]+[0]+[0]+[0],[0]+[0]+[0]+[0],[1]+[0]+[0]+[0],[0]+[0]+[0]+[0],[0]+[0]+[0]+[0]) =([1],[0],[1],[0],[0]) <Step S2> Aggregate value calculation unit 12 a ,…,12 A The flag generating unit 11 a ,…,11 A The flag generated in the step 12 is input to the aggregate value calculation unit 12. a ,…,12 A Input data is input to . Aggregate value calculation unit 12 a ,…,12 A uses the generated flags and the input data to calculate aggregate values ​​corresponding to each record (step S2). For example, when calculating the aggregated value of the sum, the aggregated value calculation unit 12 a ,…,12 A performs a secret calculation of the dot product of the attribute value and the flag for each record. For example, if [d[1]]=([0],[1],[1],[0],[0]), [a[1]]=([2],[3],[4],[1],[3]), the aggregated value calculation unit 12 a ,…,12 A calculates the dot product of [d[1]] and [a[1]] to obtain the secret value [7] as the aggregate value. For example, when calculating the aggregate value of the maximum value, the aggregate value calculation unit 12 a ,…,12 A For each record, if the secret value of the flag corresponding to the attribute value indicates that the attribute value should not be selected, a secure calculation is performed to replace the attribute value with a secret value that is sufficiently small and is a predetermined value, and then a secure calculation is performed to obtain the maximum value. For example, the confidential value of the flag [d[i,j]]=[1] indicates that the attribute value a[i,j] is to be selected, and the confidential value of the flag [d[i,j]]=[0] indicates that the attribute value a[i,j] is not to be selected. When the confidential value of the flag corresponding to an attribute value indicates that the attribute value is to be selected, the attribute value is not replaced with another value. For example, when [d[1]]=([0],[1],[1],[0],[0]), [a[1]]=([2],[3],[4],[1],[3]), and the predetermined secret value is sufficiently small=[−1000], the aggregate value calculation unit 12 a ,…,12 A teeth, By setting the secret value of the attribute value whose corresponding flag secret value element is [0] to [-1000], we first obtain ([-1000], [3], [4], [-1000], [-1000]), and then perform a secret calculation to obtain the maximum value, and obtain the secret value [4] as the aggregate value of the maximum values. For example, when calculating the aggregate value of the minimum value, the aggregate value calculation unit 12 a ,…,12 A For each record, if the secret value of the flag corresponding to the attribute value indicates that the attribute value should not be selected, a secure calculation is performed to replace the attribute value with a sufficiently large predetermined secret value, and then a secure calculation is performed to obtain the minimum value. For example, if [d[1]]=([0],[1],[1],[0],[0]), [a[1]]=([2],[3],[4],[1],[3]), a sufficiently large secret value =

[1000] If so, the aggregate value calculation unit 12 a ,…,12 A teeth, By setting the secret value of the attribute value whose corresponding flag secret value element is [0] to

[1000] , we first obtain (

[1000] , [3], [4],

[1000] ,

[1000] ), and then perform a secret calculation to obtain the minimum value, and obtain the secret value [3] as the aggregate value of the minimum value. By calculating the aggregate value using the secret value [f] of the grouping information f and the secret value of the flag, which is information on which attribute value to use, the grouping information and the information on which attribute value to use are both stored in each secret aggregate value calculation device 1. a The aggregate value can be calculated using only the same random k attribute values ​​for each group while keeping the k attribute values ​​secret (a=1, …, A). [Variations] The specific configurations of the embodiments of the disclosed technology are not limited to the configurations described above. The specific configurations of the embodiments of the disclosed technology can be appropriately modified in design without departing from the spirit of the embodiments of the disclosed technology. The various processes described in the embodiments of the disclosed technology may not only be performed chronologically in the order described, but may also be performed in parallel or individually depending on the processing capabilities of the device performing the processes or as necessary. For example, the secret aggregate value calculation device 1 a Data exchange between these components may be performed directly or via a storage unit (not shown). [Programs, recording media] The processing of each unit of each of the above-mentioned devices may be realized by a computer, in which case the processing contents of the functions that each device should have are described by a program. Then, by loading this program into the storage unit 1020 of the computer 1000 shown in Fig. 4 and operating the arithmetic processing unit 1010, the input unit 1030, the output unit 1040, the display unit 1060, etc., various processing functions of each of the above-mentioned devices are realized on the computer. The program describing the processing contents can be recorded on a computer-readable recording medium. The computer-readable recording medium is, for example, a non-transitory recording medium, specifically, a magnetic recording device, an optical disk, etc. The program may be distributed, for example, by selling, transferring, lending, etc. portable recording media such as DVDs and CD-ROMs on which the program is recorded. Furthermore, the program may be distributed by storing the program in a storage device of a server computer and transferring the program from the server computer to other computers via a network. A computer that executes such a program, for example, first stores a program recorded on a portable recording medium or a program transferred from a server computer in the auxiliary recording unit 1050, which is a non-transient storage device of the computer. Then, when executing a process, the computer reads the program stored in the auxiliary recording unit 1050, which is a non-transient storage device of the computer, into the storage unit 1020, and executes a process according to the read program. As another execution form of the program, the computer may directly read the program from the portable recording medium into the storage unit 1020 and execute a process according to the program, or may execute a process according to the received program each time a program is transferred from the server computer to the computer. Alternatively, the server computer may not transfer the program to the computer, but may execute the above-mentioned process by a so-called ASP (Application Service Provider) type service that realizes a processing function only by issuing an execution instruction and obtaining a result. Note that the program in this embodiment includes information used for processing by an electronic computer that is equivalent to a program (data that is not a direct instruction to the computer but has a nature that specifies the processing of the computer, etc.). In this embodiment, the device is configured by executing a predetermined program on a computer, but at least a part of the processing contents may be realized by hardware. For example, the flag generating unit 11 a , temporary flag generation unit 111 a , selection unit 112 a , replication section 113 a and the aggregate value calculation unit 12 amay be comprised of a processing circuit. Needless to say, other modifications can be made without departing from the spirit of the present invention. All publications, patent applications, and standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or standard was specifically and individually indicated to be incorporated by reference.

Claims

1. A secret aggregate value calculation system including a plurality of secret aggregate value calculation devices, wherein n records as input data belong to any one of a plurality of groups, each record includes secret values ​​of m attribute values ​​respectively corresponding to m attributes, [f] is a secret value of grouping information which is information about the group to which each record belongs, and a flag is information indicating which k secret values ​​of the m attribute values ​​included in the records are to be used to calculate an aggregate value, the secret value of the flag being common to each record included in each of a plurality of groups, using [f]; and an aggregate value calculation unit which calculates an aggregate value corresponding to each record using the generated flag and the input data.

2. A secret aggregate value calculation system according to claim 1, wherein the flag generation unit includes: a temporary flag generation unit that generates, for each record, a secret value of a temporary flag, which is information indicating which k secret values ​​of attribute values ​​among the m secret values ​​of attribute values ​​are to be used to calculate an aggregate value; a selection unit that selects one secret value of a temporary flag from the secret values ​​of the temporary flags of records included in each of the multiple groups; and a duplication unit that sets the secret value of the selected temporary flag as the secret value of a flag of each record included in a group that includes a record corresponding to the temporary flag.

3. A secret aggregate value calculation system according to claim 2, wherein the duplication unit sets the secret values ​​of the temporary flags of records other than the record corresponding to the secret value of the selected one temporary flag to a secret value of 0, and sets a flag whose sum of the i-th secret values ​​of the temporary flags within a group as the i-th secret value as a common flag for each record included in the group.

4. A secret aggregate value calculation device for the secret aggregate value calculation system of claim 1.

5. A secret aggregate value calculation method including a plurality of secret aggregate value calculation devices, wherein n records as input data belong to any one of a plurality of groups, each record includes secret values ​​of m attribute values ​​respectively corresponding to m attributes, and [f] is a secret value of grouping information which is information about the group to which each record belongs, and a flag generation step in which a flag generation unit generates a secret value of a flag common to each record included in each of a plurality of groups using [f], the secret value being information indicating which k secret values ​​of the m attribute values ​​included in the records are to be used to calculate an aggregate value; and an aggregate value calculation step in which an aggregate value calculation unit calculates an aggregate value corresponding to each record using the generated flag and the input data.

6. A program for causing a computer to function as each part of the secret aggregate value calculation device of claim 4.

Citation Information

Patent Citations

  • Secure aggregate maximum value system, secure aggregate minimum value system, secure computation device, secure aggregate maximum value method, secure aggregate minimum value method, and program

    WO2019208485A1