Fuzzy address matching method and device based on privacy protection

By employing a fuzzy address matching method based on word embedding and subsampling, the problem of fuzzy matching caused by differences in address information in traditional PSI technology is solved, achieving efficient and privacy-preserving merchant address recognition.

CN118916406APending Publication Date: 2024-11-08LANXIANG ZHILIAN (HANGZHOU) TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411069968.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-06
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Traditional hidden set intersection (PSI) technology cannot effectively handle fuzzy matching problems caused by differences in address information collection, and cannot identify different address representations of the same merchant.

Method used

A privacy-preserving fuzzy address matching method is adopted. After binarizing the address information through word embedding and subsampling techniques, a random vector is generated and AND logic operation is performed to filter out the sample set of address information. Finally, the common merchants are identified through the PSI set intersection protocol.

Benefits of technology

It enables efficient identification of shared merchant address information by both parties while protecting their privacy, avoiding the leakage of non-shared information and improving the accuracy and efficiency of address matching.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118916406A_ABST
    Figure CN118916406A_ABST
Patent Text Reader

Abstract

The invention discloses a fuzzy address matching method and device based on privacy protection, and the method comprises the steps: S1, obtaining first address information held by a first data holder and second address information held by a second data holder; s2, receiving a subsample set of the first address information obtained by carrying out subsampling processing on the first address information by the first data holding party, and a subsample set of the second address information obtained by carrying out subsampling processing on the second address information by the second data holding party; s3, receiving a processing result that the first data holder and the second data holder operate a PSI set intersection protocol on each subset in the subsample set of the first address information and the subsample set of the second address information; and S4, respectively sending the processing result to the first data holder and the second data holder.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a fuzzy address matching method and device based on privacy protection. Background Art

[0002] Supplier A holds some merchant address information, and supplier B also holds some merchant address information. Both parties need to identify the merchants they have in common, and then carry out some marketing activities targeting the merchants they have in common.

[0003] Supplier A and Supplier B want to obtain information about shared merchants, but they cannot expose information about non-shared merchants to each other. Traditional hidden set intersection (PSI) technology cannot solve this problem, because PSI is a precise intersection. When the information collected by both parties on a certain merchant is exactly the same, they are considered to be shared merchants. But in fact, the address information of both parties is collected independently and has certain differences. For example, for the same store - "Wanfu Convenience Store", the information collected by A is: XX Convenience Store (XX Building Store), No. 1 XXX Street, while the information collected by B is: XX Convenience Store, 1st Floor, XXX Building.

[0004] In the traditional hidden set intersection (PSI) technology, since A and B collect different information, XX convenience store is not considered to be a common merchant, but in fact XX convenience store is the same store.

[0005] In order to solve the problem that PSI cannot achieve fuzzy matching, the present invention proposes a fuzzy address matching method and device based on privacy protection. Summary of the invention

[0006] The purpose of the present invention is to provide a fuzzy address matching method and device based on privacy protection in view of the defects of the prior art.

[0007] In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0008] A fuzzy address matching method based on privacy protection, comprising:

[0009] S1. Obtaining first address information held by the first data holder and second address information held by the second data holder;

[0010] S2. Receive a sub-sample set of first address information obtained by the first data holder performing sub-sampling processing on the first address information, and a sub-sample set of second address information obtained by the second data holder performing sub-sampling processing on the second address information;

[0011] S3. Receive the processing result of the first data holder and the second data holder running the PSI set intersection protocol for each subset in the first address information subset set and the second address information subset set;

[0012] S4. Send the processing results to the first data holder and the second data holder respectively.

[0013] Furthermore, the step S1 also includes binarizing the first address information and the second address information based on the large language model to obtain the binarized first address information and the binarized second address information.

[0014] Furthermore, in step S2, the calculation method of the sub-sample set of the first address information and the sub-sample set of the second address information is:

[0015] S21. Randomly generate multiple 1024-dimensional 0-1 random vectors;

[0016] S22. Perform an AND logic operation on the binary address information and the random vector in step S21 to obtain multiple sampling results;

[0017] S23. Form a sub-sample set of address information based on multiple sampling results.

[0018] Furthermore, the processing result of running the PSI set intersection protocol in step S3 is specifically: screening the sub-samples that are the same as the sub-sample set of the first address information and the sub-sample set of the second address information, and determining whether the number of the screened identical sub-samples is greater than or equal to a preset threshold; if so, it indicates that the first address information and the second address information are the same address information.

[0019] Furthermore, after determining in step S3 that the first address information and the second address information are the same address information, the step further includes: the first data holder puts the first address information into a first set, and the second data holder puts the second address information into a second set.

[0020] Furthermore, the step S4 is specifically: sending the first set to the first data holder, and sending the second set to the second data holder.

[0021] Furthermore, a fuzzy address matching device based on privacy protection is provided, which is used to perform a fuzzy address matching method based on privacy protection. The address matching device includes a processor, and the processor is provided with:

[0022] An acquisition module, used to acquire first address information held by a first data holder and second address information held by a second data holder;

[0023] A first receiving module, configured to receive a sub-sample set of first address information obtained by a first data holder performing sub-sampling processing on the first address information, and a sub-sample set of second address information obtained by a second data holder performing sub-sampling processing on the second address information;

[0024] A second receiving module is used to receive a processing result of the first data holder and the second data holder running the PSI set intersection protocol on each subset in the first address information subset set and the second address information subset set;

[0025] The sending module is used to send the processing results to the first data holder and the second data holder respectively.

[0026] Furthermore, the processor is connected to a memory for storing data or instructions.

[0027] Furthermore, the processor is connected to a transmission device for transmitting data with an external device.

[0028] Furthermore, the processor is connected to an input / output device for inputting or outputting information.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] 1. Solved the problem that traditional PSI technology cannot intersect address information collected with differences

[0031] 2. It uses word embedding and subsampling technology instead of other complex cryptographic technologies, which is highly efficient.

[0032] 3. Both parties only obtain the intersection of all address sub-samples and the intersection of addresses, and do not obtain address information that is not in the intersection, thereby protecting the privacy of both parties. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a flow chart of a fuzzy address matching method based on privacy protection provided in Example 1. DETAILED DESCRIPTION

[0034] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0035] The purpose of the present invention is to provide a fuzzy address matching method and device based on privacy protection in view of the defects of the prior art.

[0036] Embodiment 1

[0037] This embodiment provides a fuzzy address matching method based on privacy protection. Figure 1 As shown, including:

[0038] S1. Obtaining first address information held by the first data holder and second address information held by the second data holder;

[0039] S2. Receive a sub-sample set of first address information obtained by the first data holder performing sub-sampling processing on the first address information, and a sub-sample set of second address information obtained by the second data holder performing sub-sampling processing on the second address information;

[0040] S3. Receive the processing result of the first data holder and the second data holder running the PSI set intersection protocol for each subset in the first address information subset set and the second address information subset set;

[0041] S4. Send the processing results to the first data holder and the second data holder respectively.

[0042] This embodiment includes a first kitchenware holder and a second data holder. The first data holder is set as Party A and the second data holder is set as Party B.

[0043] Party A holds the first address information DA={a1,a2,...,a n}, Party B holds the second address information DB = {b1, b2, ..., b m}. Party A and Party B both preset the Hamming weight of the random vector as w, the number of sub-sampling as s, and the preset threshold as t.

[0044] In step S1, first address information held by a first data holder and second address information held by a second data holder are obtained.

[0045] Party A holds the first address information DA={a1,a2,...,a n}Use the word embedding algorithm in the large language model quentinz / bge-large-zh-v1.5 to process, specifically: for all text address information a1, a2, ..., a n Encode them one by one and encode them into vectors: u1,u2,...,u n , each vector u i =(u i1 ,u i2,...,u i1024 The dimension of )(i=1,...,n) is 1024, and the elements of each dimension are signed floating-point numbers.

[0046] In this embodiment, for each vector u i Perform the following transformation: convert the floating point vector u i Binarize and convert to 0-1 vector. The conversion to 0-1 vector is as follows: i Calculate the mean value of all elements in ui ; for vector u i Each element in is judged one by one, if it is greater than the average value mean ui , then set the element to 1 if it is less than or equal to the mean value mean ui , then set the element to 0.

[0047] According to the above method, Party A obtains the binary first address information Binary DA ={u1,u2,...,u n}.

[0048] Party B holds the second address information DB = {b1, b2, ..., b m}Use the word embedding algorithm in the large language model quentinz / bge-large-zh-v1.5 to process, specifically: for all text address information b1, b2, ..., b m Encode them one by one and encode them into vectors: v1,v2,...,v m , each vector v j =(v j1 ,v j2 ,...,v j1024 The dimension of )(j=1,...,m) is 1024, and the elements of each dimension are signed floating-point numbers.

[0049] In this embodiment, for each vector v j Perform the following conversion: convert the floating point vector v j Binarize and convert to 0-1 vector. The conversion to 0-1 vector is as follows: j Calculate the mean value of all elements in vj ; for vector v i Each element in is judged one by one, if it is greater than the average value mean vj , then set the element to 1 if it is less than or equal to the mean value mean vj , then set the element to 0.

[0050] According to the above method, Party B obtains the second binary address information Binary DB ={v1,v2,...,v m}.

[0051] In step S2, a sub-sample set of first address information obtained by sub-sampling the first address information by the first data holder and a sub-sample set of second address information obtained by sub-sampling the second address information by the second data holder are received.

[0052] A's first binary address information DA ={u1,u2,...,u n Every information u i =(u i1 ,u i2 ,...,u i1024 )(i=1,...,n) performs sub-sampling processing, specifically:

[0053] A1. Randomly generate multiple 1024-dimensional 0-1 random vectors.

[0054] Randomly generate s 1024-dimensional 0-1 random vectors, rand vec1 ,rand vec2 ,...,rand vecs The Hamming weight of these random vectors is less than w;

[0055] A2. Perform AND logic operations on the binary address information and the random vectors in step A1 to obtain multiple sampling results.

[0056] Party A will i Respectively with rand vec1 ,rand vec2 ,...,rand vecs Perform AND logic operation to obtain s sampling results, u i _sample1,u i _sample2,...,u i _sample s .

[0057] A3. Forming a sub-sample set of address information based on multiple sampling results

[0058] Party A has a DA ={u1,u2,...,u n Each piece of information of} can get the following sub-sample set:

[0059] Sample DA={u1_sample={u1_sample1,u1_sample2,...,u1_sample s},u2_sample={u

[0060] 2_sample1,u2_sample2,...,u2_sample s},..,u n _sample={u n _sample1,u n _sample2,...,u n _sample s}}; Among them, the sub-sample set of party A contains n subsets, and each subset has s elements.

[0061] Party B's second binary address information DB ={v1,v2,...,v m}Every information v j =(v j1 ,v j2 ,...,v j1024 )(j=1,...,m) for sub-sampling, specifically:

[0062] B1. Randomly generate multiple 1024-dimensional 0-1 random vectors.

[0063] Randomly generate s 1024-dimensional 0-1 random vectors, rand vec1 ,rand vec2 ,...,rand vecs The Hamming weight of these random vectors is less than w;

[0064] B2. Perform AND logic operations on the binary address information and the random vector in step B1 to obtain multiple sampling results.

[0065] Party B will v j Respectively with rand vec1 ,rand vec2 ,...,rand vecs Perform AND logic operation to obtain s sampling results, v j _sample1,v j _sample2,...,v j _sample s .

[0066] B3. Forming a sub-sample set of address information based on multiple sampling results

[0067] Party B has a DB ={v1,v2,...,v m Each piece of information of} can get the following sub-sample set:

[0068] Sample DB ={v1_sample={v1_sample1,v1_sample2,...,v1_sample s},v2_sample={v

[0069] 2_sample1,v2_sample2,...,v2_sample s},..,v m _sample={v m _sample1,v m _sample2,...,v m _sample s}}; Among them, the sub-sample set of party B contains m subsets, and each subset has s elements.

[0070] In step S3, the processing result of the first data holder and the second data holder running the PSI set intersection protocol on each subset of the first address information subset set and the second address information subset set is received.

[0071] Party A and Party B target Sample DA and Sample DB Each subset in runs the traditional PSI set intersection protocol as follows, the principle of which is: for the first address information and the second address information, the s sub-samples in the sub-sample set of the first address information and the s sub-samples in the sub-sample set of the second address information are screened for the same sub-samples, and it is determined whether the number of the screened same sub-samples is greater than or equal to a preset threshold t. If so, it means that the first address information and the second address information are the same address information, specifically:

[0072] A Let the set S_A = {}

[0073] B Let the set S_B = {}

[0074] for i=1,2,...,n

[0075] for j=1,2,...m

[0076] sample_intersection_set=PSI(u_i_sample,v_j_sample)

[0077] If the number of elements in sample_intersection_set is greater than or equal to t:

[0078] Then party A will take element u i Put it into the set SA,

[0079] else:

[0080] Then B will take the element v j Put it into the set SB.

[0081] In step S4, the processing results are sent to the first data holder and the second data holder respectively.

[0082] Party A obtains set SA, and Party B obtains set SB. The elements in set SA and set SB are the information of merchants shared by both parties.

[0083] This embodiment has the following beneficial effects:

[0084] 1. Solved the problem that traditional PSI technology cannot intersect address information collected with differences

[0085] 2. It uses word embedding and subsampling technology instead of other complex cryptographic technologies, which is highly efficient.

[0086] 3. Both parties only obtain the intersection of all address sub-samples and the intersection of addresses, and do not obtain address information that is not in the intersection, thereby protecting the privacy of both parties.

[0087] Embodiment 2

[0088] This embodiment further provides a fuzzy address matching device based on privacy protection, which is used to execute a fuzzy address matching method based on privacy protection in embodiment 1. The address matching device includes a processor, and the processor is provided with:

[0089] An acquisition module, used to acquire first address information held by a first data holder and second address information held by a second data holder;

[0090] A first receiving module, configured to receive a sub-sample set of first address information obtained by a first data holder performing sub-sampling processing on the first address information, and a sub-sample set of second address information obtained by a second data holder performing sub-sampling processing on the second address information;

[0091] A second receiving module is used to receive a processing result of the first data holder and the second data holder running the PSI set intersection protocol on each subset in the first address information subset set and the second address information subset set;

[0092] The sending module is used to send the processing results to the first data holder and the second data holder respectively.

[0093] The processor described in this embodiment may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0094] The processor described in the present embodiment is connected with a memory for storing data or instructions. The memory may include a large capacity memory for data or instructions. For example, but not limitation, the memory may include a hard disk drive (HDD), a floppy disk drive, a solid state drive (SSD), a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory may be inside or outside the data processing device. In a specific embodiment, the memory is a non-volatile memory. In a specific embodiment, the memory includes a read-only memory (ROM) and a random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (Programmable Read-Only Memory, PROM for short), an erasable PROM (Erasable Programmable Read-Only Memory, EPROM for short), an electrically erasable PROM (Electrically Erasable Programmable Read-Only Memory, EEPROM for short), an electrically alterable ROM (Electrically Alterable Read-Only Memory, EAROM for short) or a flash memory (FLASH) or a combination of two or more of these. Under appropriate circumstances, the RAM can be a static random access memory (SRAM) or a dynamic random access memory (DRAM), wherein the DRAM can be a fast page mode dynamic random access memory (FPMDRAM), an extended data output dynamic random access memory (EDODRAM), a synchronous dynamic random access memory (SDRAM), etc.

[0095] The memory can be used to store or cache various data files that need to be processed and / or used for communication, as well as possible computer program instructions executed by the processor. The processor reads and executes the computer program instructions stored in the memory.

[0096] The processor described in this embodiment is connected to a transmission device for transmitting data with an external device. The transmission device can be used to receive or send data via a network. The above-mentioned specific examples of the network may include a wired or wireless network provided by a communication provider of the electronic device. In one example, the transmission device includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device can be a radio frequency (Radio Frequency, referred to as RF) module, which is used to communicate with the Internet wirelessly.

[0097] The processor of this embodiment is connected with an input / output device for inputting or outputting information. The input / output device is used for inputting or outputting information. In this embodiment, the input information is the address information of both parties, and the output information is the information of the merchant shared by both parties.

[0098] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A fuzzy address matching method based on privacy protection, characterized in that: include: S1. Obtaining first address information held by the first data holder and second address information held by the second data holder; S2. Receive a sub-sample set of first address information obtained by the first data holder performing sub-sampling processing on the first address information, and a sub-sample set of second address information obtained by the second data holder performing sub-sampling processing on the second address information; S3. Receive the processing result of the first data holder and the second data holder running the PSI set intersection protocol for each subset in the first address information subset set and the second address information subset set; S4. Send the processing results to the first data holder and the second data holder respectively.

2. A privacy protection based fuzzy address matching method according to claim 1, characterized in that: The step S1 also includes binarizing the first address information and the second address information based on the large language model to obtain the binarized first address information and the binarized second address information.

3. A privacy protection based fuzzy address matching method according to claim 2, characterized in that: In step S2, the calculation method of the sub-sample set of the first address information and the sub-sample set of the second address information is: S21. Randomly generate multiple 1024-dimensional 0-1 random vectors; S22. Perform an AND logic operation on the binary address information and the random vector in step S21 to obtain multiple sampling results; S23. Form a sub-sample set of address information based on multiple sampling results.

4. A privacy protection based fuzzy address matching method according to claim 3, characterized in that: The processing result of running the PSI set intersection protocol in step S3 is specifically: screening the sub-samples of the first address information sub-sample set that are the same as the sub-sample set of the second address information, and determining whether the number of the screened identical sub-samples is greater than or equal to a preset threshold. If so, it indicates that the first address information and the second address information are the same address information.

5. A privacy protection based fuzzy address matching method according to claim 4, characterized in that: After determining in step S3 that the first address information and the second address information are the same address information, the method further includes: the first data holder puts the first address information into a first set, and the second data holder puts the second address information into a second set.

6. A privacy protection based fuzzy address matching method according to claim 5, characterized in that: The step S4 specifically includes: sending the first set to the first data holder, and sending the second set to the second data holder.

7. A fuzzy address matching device based on privacy protection, used to execute a fuzzy address matching method based on privacy protection according to any one of claims 1 to 6, the address matching device comprising a processor, characterized in that: The processor is provided with: An acquisition module, used to acquire first address information held by a first data holder and second address information held by a second data holder; A first receiving module, configured to receive a sub-sample set of first address information obtained by a first data holder performing sub-sampling processing on the first address information, and a sub-sample set of second address information obtained by a second data holder performing sub-sampling processing on the second address information; A second receiving module is used to receive a processing result of the first data holder and the second data holder running the PSI set intersection protocol on each subset in the first address information subset set and the second address information subset set; The sending module is used to send the processing results to the first data holder and the second data holder respectively.

8. The fuzzy address matching device based on privacy protection according to claim 7, characterized in that: The processor is connected to a memory for storing data or instructions.

9. The fuzzy address matching device based on privacy protection according to claim 7, characterized in that: The processor is connected with a transmission device for transmitting data with an external device.

10. The fuzzy address matching device based on privacy protection according to claim 7, characterized in that: The processor is connected to an input / output device for inputting or outputting information.