CN105592085B - The method for secret protection of facing position perception recommender system - Google Patents
The method for secret protection of facing position perception recommender system Download PDFInfo
- Publication number
- CN105592085B CN105592085B CN201510979917.5A CN201510979917A CN105592085B CN 105592085 B CN105592085 B CN 105592085B CN 201510979917 A CN201510979917 A CN 201510979917A CN 105592085 B CN105592085 B CN 105592085B
- Authority
- CN
- China
- Prior art keywords
- encryption
- information
- recommended user
- isp
- cloud computing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/52—Network services specially adapted for the location of the user terminal
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/04—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks
- H04L63/0428—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks wherein the data content is protected, e.g. by encrypting or encapsulating the payload
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/2866—Architectures; Arrangements
- H04L67/30—Profiles
- H04L67/306—User profiles
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/53—Network services using third party service providers
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/55—Push-based network services
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/56—Provisioning of proxy services
- H04L67/566—Grouping or aggregating service requests, e.g. for unified processing
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Security & Cryptography (AREA)
- Computer Hardware Design (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The method for secret protection for the location aware recommender system based on cryptography that the invention discloses a kind of mainly solves the problem of that existing system can bring information loss to reduce recommendation service quality while protecting recommended user's privacy.Its technical solution is: first storing all historical datas in cloud computing platform in the form of ciphertext；ISP calculates the similarity between each position point that recommended user accessed using Paillier Encryption Algorithm；Recommended user and ISP calculate the recommendation results for meeting user's requirement using Paillier Encryption Algorithm, El-Gamal Encryption Algorithm and comparable Encryption Algorithm.The result showing method of emulation testing can realize recommendation service rapidly and efficiently under conditions of ensure that the safety of the private data of private data and service provider of recommended user.The secret protection that can be used in the outer packet system of data under cloud computing environment.
Description
Technical field
The invention belongs to information security fields, are related to a kind of method for secret protection, can be used under cloud computing environment outside data
Secret protection in packet system.
Background technique
As city calculates the fast development with GPS device, location based service LBS has been widely used in people's
In daily life, more conveniences are provided for our life.Feel for example, Foursquare allows recommended user to retrieve oneself
The location point POI of interest, and recommended user is allowed to evaluate when leaving service obtained.However, with various each
How rapidly and efficiently, accurately the appearance of the service of sample, a large amount of LBS service are full of in our life, therefore, obtain
The service that we want becomes the problem of recommended user's urgent need to resolve.The appearance of recommender system is exactly to the complete of problem as described above
Beauty solves.Recommender system can carry out recommendation prediction according to the hobby of recommended user and the evaluation information of recommended user, and will
Prediction result feeds back to service requester.However, the recommender system based on location aware is not only compared with traditional recommender system
The evaluation information of recommended user is considered, it is also contemplated that the time space position information of recommended user.Therefore, pushing away based on location aware
The system of recommending can provide the recommendation service in specified region for recommended user.
It is safely and efficiently used currently, the main challenge of recommender system is how to realize in a large amount of location point for recommendation
Recommended at family.However, traditional recommended technology not can be used directly the recommendation in LBS service, reason mainly has following two o'clock:
Firstly, since Spatio-temporal Data is multifarious and is that fast-changing, traditional recommender system needs to consume largely
Resource stores and calculates these data, and therefore, a large amount of LBS service supplier starts to realize data by cloud computing service
Storage and calculating.By the way that by recommended user's data and recommendation computation migration to cloud computing platform, ISP can protected
Card recommends the consumption that own resource is substantially reduced under the premise of quality.For example, Netflix closes it certainly in part of in August, 2015
The last one data center that body possesses stores all Data Migrations to Amazon cloud computing platform.However, will count
Other Privacy Protections will be brought again according to third party cloud computing platform is moved to.Due to recommended user submit data and push away
Recommending result includes many privacy informations, for example, location information, preference information etc., cloud computing platform server can pass through these
Which position when which recommended user of information inference be located at.Therefore, cloud computing platform may track certain recommendations and use
The preference information of recommended user is sent to other attackers by family.If service provider not can be well protected recommendation and use
The privacy information at family, recommended user may refuse the recommendation service provided using service provider because of concern about disclosure privacy,
So Privacy Protection is another big factors for hindering recommender system service development.
In view of factors above, protecting the privacy of recommended user is developed based on location aware recommender system.However,
The protection of data not only only includes the privacy information of recommended user, further includes recommended user's history that service provider is collected into
Data information.For service provider as a commercial company, the data being collected into may be considered its privately owned assets, Bu Nengwu
The external disclosure repaid, therefore, when the data that service provider is possessed are sent to cloud computing platform, it is necessary to data processing
It uploads again afterwards, so that cloud computing platform can not obtain its raw information.
Existing recommendation service secret protection, such as: " framework of the intimacy protection system for recommendation service " patent application
(application number: 201380031170.X) and " based on correlation rule meet recommended user's secret protection personalized recommendation method and
System " (application number: 201410283430.9)) mostly uses anonymity scheme or difference privacy mechanism, therefore, these systems exist
Certain information loss is also brought along while protecting recommended user's privacy, to reduce recommendation service quality.
Data outsourcing is carried out in half believable cloud computing environment, necessarily system designs the key realized for safety,
Meanwhile how under the premise of guaranteeing safety, the recommendation service that high quality can be provided for recommended user is also location aware
The key point of recommender system design.
Summary of the invention
The technical problem to be solved by the present invention is to for the privacy concern in location aware recommender system, propose it is a kind of towards
The method for secret protection of location aware recommender system, to guarantee the correctness of recommended user's personal secrets and recommended user's information,
Improve recommendation service quality.
Technical thought of the invention is: for the diversity and dynamic changeability of position related data, proposing to recommend system
Beyond the clouds by the history data store of recommended user, service provider only needs periodically to collect recommended user's system service provider
Historical data, and it is uploaded to third party cloud computing platform after encryption；For the safety problem after data outsourcing and recommend to use
The privacy concern at family proposes the Privacy Preservation Mechanism based on cryptography theory, so that service provider is guaranteeing that recommended user is hidden
Recommendation service is provided under the premise of private safety for recommended user.Its implementation is as follows:
The method for secret protection of facing position perception recommender system, it is characterised in that by cryptography theory and data outsourcing
Computational theory is applied in recommender system, and step includes:
(1) ISP is extended the attribute information of recommended user's location point, and to the attribute information after extension
It is encrypted, then encrypted information is sent to cloud computing platform and is stored；
(2) ISP periodically collects the history evaluation information of recommended user, and to the history evaluation information being collected into
Cloud computing platform is sent to after encrypting using Paillier Encryption Algorithm；
(3) it after cloud computing platform receives the history evaluation information that ISP sends, is integrated into and has stored in
In the data set in cloud, then aminated polyepichlorohydrin is carried out to the data set, and polymerization result is sent back into ISP；
(4) ISP decrypts polymerization result and calculates the similarity between each location point, then to similarity matrix
After being converted, it is sent to cloud computing platform and is stored；
(5) recommended user directly sends to cloud computing platform with the comparable encrypted area-of-interest of Encryption Algorithm, cloud
After computing platform receives data, the location point met in recommended user's area-of-interest is filtered out using comparable Encryption Algorithm；
(6) cloud computing platform obtains bilayer again to the attribute information commutative encryption algorithm for encryption for filtering out location point
Encryption as a result, simultaneously, extract the recommended user history evaluation information and corresponding similarity information, and carry out polymerization fortune
It calculates, then the result of double layer encryption and polymerization result is sent to ISP；
(7) ISP receive cloud computing platform transmission double layer encryption result after, it is calculated with commutative decryption
Method decrypts internal layer encryption, obtains the attribute information of recommended user's key encryption, meanwhile, the polymerization result of cloud computing platform is used
Paillier algorithm is decrypted and is calculated, and obtains predictive information, and the attribute information and predictive information are sent to recommendation and used
Family；
(8) recommended user decrypts the commutative decipherment algorithm of the attribute information received, obtains attribute information in plain text, simultaneously
The predictive information received is calculated, prediction and evaluation information is obtained in plain text, successively selects the highest k position of prediction and evaluation information
It sets a little, and shows this k location point on recommended user's area-of-interest according to its attribute information.
The present invention has the advantage that
1. ISP's Paillier algorithm for encryption history evaluation information, so that the history evaluation information of recommended user
It is calculated on cloud computing platform with ciphertext form, ensure that the safety of history evaluation information on cloud computing platform.
2. ISP is encrypted with attribute information of the commutative encryption algorithm to location point, with comparable Encryption Algorithm pair
The attribute information encryption expanded, so that the attribute information of location point is calculated on cloud computing platform with ciphertext form, ensure that
The safety of the attribute information of location point on cloud computing platform.
3. recommended user encrypts its interested region with comparable Encryption Algorithm, cloud computing platform commutative encryption
Algorithm encrypts location point attribute information, so that the location information of recommended user and prediction recommendation results are being taken in the form of ciphertext
It is calculated on business supplier and cloud computing platform, ensure that prediction recommendation results on ISP and cloud computing platform
Data safety.
To sum up, this invention ensures that the safety of the private data of the private data and service provider of recommended user, is realized
The secret protection of facing position perception recommender system.
Detailed description of the invention
Fig. 1 is the location aware recommender system illustraton of model that the present invention uses；
Fig. 2 is implementation flow chart of the invention；
Fig. 3 is the schematic diagram of history evaluation information fusion during calculating similarity in the present invention；
Fig. 4 is recommendation quality simulation assessment figure of the invention.
Specific embodiment
Below in conjunction with attached drawing, the present invention will be described in detail:
Referring to Fig.1, the location aware recommender system service model with privacy protection function that the present invention uses, by
Trusted party, ISP, cloud computing platform and recommended user are constituted.Trusted party is mainly responsible for generation key, and is
Recommended user and ISP distribute key, and within the system, trusted party is unique believable entity completely；Service provides
Person is mainly that recommended user provides recommendation service, and ISP can periodically collect the history evaluation information of recommended user, and will
The information being collected into is sent to cloud computing platform storage, and within the system, ISP has limited storage resource and meter
Therefore cloud computing platform can be contracted out to for the information being collected into and store and process by calculating resource, meanwhile, ISP
It is half believable, can correctly calculates recommendation results for recommended user, but " can also peep " privacy of recommended user；Cloud meter
Calculating platform is main storage and calculating center in system, possesses unlimited storage and computing capability, is mainly responsible for secure storage
The information that ISP uploads, and Auxiliary Service Provider is that recommended user carries out recommendation calculating, within the system, cloud computing
Platform is also half believable, other recommended users can correctly be helped to handle data, but also to the source of ISP
The privacy information of data and recommended user are interested.
The present invention is based on the protections of Fig. 1 system privacy to be divided into two stages: similarity calculation stage and recommended user test and assess in advance
Valence information phase.The similarity calculation stage is primarily referred to as the process of service provider data's outsourcing, and in this stage, cloud computing is flat
Platform is also responsible for Auxiliary Service Provider in the case where ciphertext and calculates between each position point other than carrying out secure storage to data
Similarity；Recommended user's prediction and evaluation information phase is then mainly responsible for be calculated in the case where guaranteeing recommended user's personal secrets
Meet the recommendation results of recommended user request.
Referring to Fig. 2, steps are as follows for realization of the invention:
Step 1, ISP is extended location point attribute structure body and encrypts respectively.
ISP the storage of the attribute information of each location point beyond the clouds, these attribute informations include location information,
Title code name etc..It can also restore to guarantee these attribute informations after encryption, while guarantee that encrypted position coordinates have
Comparability needs to be extended the structural body, and steps are as follows:
1a) ISP is by attribute information < I of location pointN,IL>it is extended to<< IN,IL>,IL>, wherein INIndicate each
The title code name coordinate set of location point, ILIndicate the position coordinates set of each position point；
1b) ISP is to former attribute information < IN,IL> encrypts to obtain with the commutative encryption key of ISP
CiphertextTo the position attribution < I expandedL> encrypts to obtain ciphertext Enc with that can may compare Encryption Algorithm
(IL)；
1c) ISP is by ciphertextCloud computing platform is sent to be stored.
Step 2, ISP collects history evaluation data and data is contracted out to cloud computing platform.
2a) ISP can periodically collect the history evaluation information of recommended user, generate history evaluation information matrix Rt,
And information matrix will be copied to obtain backup matrix Rt'；
2b) ISP is with Paillier Encryption Algorithm to history evaluation information matrix RtEncrypt the ciphertext generated
2c) according to Γ (m, s, α)=s* (m* α2+ c) transformation rule to backup matrix Rt' converted, after obtaining conversion
Result are as follows: At=Γ (Rt', s, α), wherein s is random number, and c is random number, and α is Big prime, and m indicates the number for needing to convert
According to AtIndicate history evaluation information transformed matrix.
2d) ISP is by ciphertextIt is sent to cloud computing platform.
Step 3, cloud computing platform carries out aminated polyepichlorohydrin to history evaluation information.
Referring to Fig. 3, this step is implemented as follows:
3a) cloud computing platform is integrated after receiving the ciphertext that ISP sends:
Firstly, will be to history evaluation information matrix RtEncrypt the ciphertext generatedIt has been deposited on cloud computing platform
The ciphertext that history evaluation information matrix R' encryption is generated of storageIntegrated, after being integrated to history evaluation
The ciphertext that information matrix R encryption generates
Then, by history evaluation information transformed matrix AtWith history evaluation information conversion stored on cloud computing platform
Matrix A ' integration, the history evaluation information transformed matrix A after being integrated；
3b) by above-mentioned ciphertextAminated polyepichlorohydrin is carried out with the history evaluation information transformed matrix A after integration, is obtained
Polymer matrix B:
Wherein, rjiIndicate i-th of recommended user to the history evaluation information of j-th of location point, akiIndicate i-th of recommendation
Conversion information of the user to the history evaluation information of k-th of location point, v expression recommended user's number,
After aminated polyepichlorohydrin, so that the data matrix for being stored in cloud computing platform is compressed, reduce to service
Supplier's storage capacity demand；
3c) polymer matrix B is sent to ISP by cloud computing platform, ISP in having enough spaces into
Row storage.
Step 4, ISP calculates the similarity between each location point.
After 4a) ISP receives the polymer matrix B that cloud computing platform is sent, with Paillier decipherment algorithm to polymerization
Matrix B is decrypted, the plaintext matrix after being polymerizeWhereinIt indicates to use
Paillier decipherment algorithm decrypts matrix B；
4b) element b' of the ISP to the plaintext matrix B' after polymerizationjkIt is converted, obtains location point j and position
The point multiplication operation result of the evaluation information of point k:
4c) to djkIt carries out operation and obtains the similarity sim between j-th of location point and k-th of location pointjk:
Wherein s-1Indicate that random number s's is inverse,Indicate the mould of all evaluation informations of j-th of location point,It indicates
The mould of all evaluation informations of k-th of location point；
With the similarity sim between j-th of location point and k-th of location pointjkConstitute similarity matrix Sim；
4d) similarity matrix Sim is converted:
Since ISP has limited computing resource, similarity matrix is sent cloud computing by the present invention
Platform end is stored, and carries out the request of customer in response.But private property of the similarity matrix as ISP,
It cannot be directly stored in cloud computing platform in the form of plaintext, so the present invention is according to Γ (m, s, α)=s* (m* α2+ c)
Transformation rule converts similarity matrix Sim, the similarity matrix F=Γ (Sim, s, α) after being converted；
4d) ISP sends cloud computing platform for the similarity matrix F after conversion and stores.
Step 5, cloud computing platform filters out location point.
5a) recommended user is using comparable Encryption Algorithm to its interested region { xu±Δx,yu± Δ y } added
It is close, obtain ciphertext Enc (xu±Δx,yu± Δ y) and Der (xu±Δx,yu± Δ y), wherein (xu,yu) indicate recommended user's
Coordinate, Δ x indicate xuVariation range, Δ y indicate yuVariation range, Enc (xu±Δx,yu± Δ y) is indicated with comparable
Encrypt the ciphertext encrypted to user's area-of-interest range, Der (xu±Δx,yu± Δ y) is indicated with comparable encryption to user
The token that the encryption of area-of-interest range generates；
5b) for recommended user to service provider registers service request, ISP sends parameter alpha to recommended user；
5c) recommended user is by above-mentioned ciphertext Enc (xu±Δx,yu± Δ y) and token Der (xu±Δx,yu± Δ y),
s′,β2,α2γ is sent to cloud computing platform, and wherein s ' is a random number, and β is a Big prime, and γ is a Big prime；
After 5c) cloud computing platform receives ciphertext and the token of recommended user, recommendation is filtered out using comparable Encryption Algorithm
Location point in user's area-of-interest, screening conditions are as follows:
Wherein param indicates the parameter of comparable Encryption Algorithm, Enc (vix) indicate flat to cloud computing using comparable encryption
The ciphertext of the abscissa encryption of the location point stored on platform, Enc (viy) indicate using comparable encryption to being deposited on cloud computing platform
The ciphertext of the ordinate encryption of the location point of storage；
5d) attribute information for filtering out location point is stored in set H.
Step 6, cloud computing platform encrypted location point attribute information and to the information of user carry out aminated polyepichlorohydrin.
6a) cloud computing platform encrypts the attribute of the location point in set H using commutative encryption algorithm again, obtains close
TextWherein I'NIndicate the title code name coordinate set of location point in set H, I'LIndicate position in set H
Position coordinates set a little is set,Indicate ISP's commutative encryption public key to location point attribute (I'N,I
'L) encryption ciphertext,Indicate the commutative encryption public key pair with recommended userEncryption
Ciphertext；
6b) cloud computing platform extracts location point and the recommended user in the history evaluation information and set H of recommended user and visits
The similarity information for the location point asked；
6c) to the similarity information summation operation of extraction, polymerization calculating is carried out to the evaluation information of extraction:
Wherein, tiIndicate the sum of all similarities, q on location point iuiIndicate the poly- of the history evaluation information of recommended user u
It closes as a result, filThe similarity between location point i and location point l after indicating conversion, zilFor random number, rlIndicate recommended user's
History evaluation information, N indicate the parameter in Paillier Encryption Algorithm, tiConstitute similarity set T, quiConstitute evaluation information collection
Close Q；
6d) cloud computing platform is by data setIt is sent to ISP.
Step 7, the predictive information containing random number is calculated in ISP.
7a) ISP receive cloud computing platform transmission double layer encryption result after, it is calculated with commutative encryption
Method decrypts internal layer encryption, obtains the encrypted attribute information of recommended user's key
7b) ISP decrypts to obtain in plain text to the evaluation information set Q received with Paillier decipherment algorithm
7c) according toReversal transformation rule, to above-mentioned set T andReversal is carried out, the predictive information containing random number is calculated
7d) ISP is by attribute informationRecommended user is sent to predictive information R'.
Step 8, recommended user successively selects the highest k location point of prediction and evaluation information.
8a) ISP is to the attribute information receivedIt is decrypted, is sieved with Paillier decipherment algorithm
The prediction and evaluation data < I' of favored areaN,I'L>；
8b) recommended user carries out reversal to the predictive information R' received, obtains its corresponding plaintext prediction and evaluation information
Rp=Γ-1(R',s',β)；
8c) recommended user successively selects the highest k location point of prediction and evaluation information, and is being recommended according to its attribute information
This k location point is shown on user's area-of-interest.
Effect of the invention can further illustrate that facing position perceives the secret protection side of recommender system by following experiment
The recommendation quality of method.
Using Foursquare, really data set tests the present invention to the present invention, which includes 2153471 and push away
User is recommended to 2809581 history evaluation information of 1143092 location points.It is tested by simulated program, obtains the present invention and push away
Recommending quality, test results are shown in figure 4.
The probability that test result shows that prediction and evaluation information error number is 0 is 78.3%, prediction and evaluation information error
Number is respectively 7.5% and 3.8% for 4 and 5 probability, illustrates that recommendation quality of the invention is sufficiently high, is able to satisfy the demand of user.
Claims (9)
1. the method for secret protection of facing position perception recommender system, it is characterised in that by cryptography theory and data outsourcing meter
It calculates theory to be applied in recommender system, step includes:
(1) ISP is extended the attribute information of recommended user's location point, and carries out to the attribute information after extension
Encryption, then encrypted information is sent to cloud computing platform and is stored；
(2) ISP periodically collects the history evaluation information of recommended user, and the history evaluation use of information to being collected into
Cloud computing platform is sent to after the encryption of Paillier Encryption Algorithm；
(3) it after cloud computing platform receives the history evaluation information that ISP sends, is integrated into and has stored in cloud
Data set in, then aminated polyepichlorohydrin is carried out to the data set, and polymerization result is sent back into ISP；
(4) ISP decrypts polymerization result and calculates the similarity between each location point, then carries out to similarity matrix
After conversion, it is sent to cloud computing platform and is stored；
(5) recommended user directly sends to cloud computing platform with the comparable encrypted area-of-interest of Encryption Algorithm, cloud computing
After platform receives data, the location point met in recommended user's area-of-interest is filtered out using comparable Encryption Algorithm；
(6) cloud computing platform obtains double layer encryption again to the attribute information commutative encryption algorithm for encryption for filtering out location point
As a result, simultaneously, extract the recommended user history evaluation information and corresponding similarity information, and carry out aminated polyepichlorohydrin, then
The result of double layer encryption and polymerization result are sent to ISP；
(7) after ISP receives the result of the double layer encryption of cloud computing platform transmission, to it with commutative decipherment algorithm solution
Close internal layer encryption, obtains the attribute information of recommended user's key encryption, meanwhile, the polymerization result of cloud computing platform is used
Paillier algorithm is decrypted and is calculated, and obtains predictive information, and the attribute information and predictive information are sent to recommendation and used
Family；
(8) recommended user decrypts the commutative decipherment algorithm of the attribute information received, obtains attribute information in plain text, while to receipts
To predictive information calculated, obtain prediction and evaluation information in plain text, successively select the highest k position of prediction and evaluation information
Point, and this k location point is shown on recommended user's area-of-interest according to its attribute information.
2. according to the method described in claim 1, wherein ISP believes the attribute of recommended user's location point in step (1)
Breath is extended, and refers to attribute information < I of the ISP by location pointN,IL> is extended to < < IN,IL>, IL>, wherein
INIndicate the title code name coordinate set of each position point, ILIndicate the position coordinates set of each position point.
3. according to the method described in claim 2, wherein ISP adds the attribute information after extension in step (1)
It is close, refer to ISP to former attribute information < IN,IL> encrypts to obtain ciphertext with the commutative encryption key of ISPTo the position attribution < I expandedL> is encrypted to obtain ciphertext Enc (I with comparable Encryption AlgorithmL)。
4. according to the method described in claim 1, the wherein history evaluation information in step (2), refer to user to accessing
The evaluation of location point.
5. according to the method described in claim 1, wherein to the history evaluation use of information Paillier being collected into step (2)
Encryption Algorithm encryption, carries out as follows:
(2.1) the history evaluation information that ISP can periodically collect user generates history evaluation information matrix Rt, to the information
Matrix is copied to obtain backup matrix Rt'；
(2.2) ISP is with Paillier Encryption Algorithm to information matrix RtEncryption generates ciphertext
(2.3) according to Γ (m, s, α)=s* (m* α2+ c) transformation rule to backup matrix Rt' converted, after being converted
As a result are as follows: At=Γ (Rt', s, α), wherein s is random number, and c is random number, and α is Big prime, and m indicates the data for needing to convert,
AtIndicate history evaluation information transformed matrix.
6. according to the method described in claim 1, wherein carrying out polymerization fortune to the history evaluation information after integration in step (3)
It calculates, carries out as follows:
Wherein B indicates polymerization result matrix,Indicate to Paillier Encryption Algorithm to information matrix RtEncryption generates
CiphertextThe ciphertext obtained after integration, A are indicated to history evaluation information transformed matrix AtThe conversion square obtained after integration
Battle array.
7. according to the method described in claim 1, wherein in step (5) recommended user to cloud platform initiate service request, then,
Cloud platform filters out the location point met in recommended user's area-of-interest using comparable Encryption Algorithm, as follows into
Row:
(5.1) recommended user is using comparable Encryption Algorithm to its interested region { xu±Δx,yu± Δ y } it is encrypted,
Obtain ciphertext Enc (xu±Δx,yu± Δ y) and Der (xu±Δx,yu± Δ y), wherein (xu,yu) indicate recommended user seat
Mark, Δ x indicate xuVariation range, Δ y indicate yuVariation range, Enc (xu±Δx,yu± Δ y) indicates to be added with comparable
The ciphertext that close algorithm encrypts area-of-interest range, Der (xu±Δx,yu± Δ y) is indicated with comparable Encryption Algorithm to sense
The token that the encryption of interest regional scope generates；
(5.2) recommended user is by above-mentioned ciphertext Enc (xu±Δx,yu± Δ y) and token Der (xu±Δx,yu± Δ y) is sent to
Cloud platform；
(5.3) after cloud platform receives ciphertext and the token of recommended user, recommended user is filtered out using comparable Encryption Algorithm and is felt
Location point in interest region, screening conditions are as follows:
And the related data of screening location point is stored in data set H, the length of data set H is h,
Wherein param indicates the parameter of comparable Encryption Algorithm, Enc (vix) indicate flat to cloud computing using comparable Encryption Algorithm
The ciphertext of the abscissa encryption of the location point stored on platform, Enc (viy) indicate using comparable Encryption Algorithm to cloud computing platform
The ciphertext of the ordinate encryption of the location point of upper storage.
8. according to the method described in claim 1, wherein in step (6) cloud computing platform to the attribute information for filtering out location point
With commutative encryption algorithm for encryption, refer to attribute information of the cloud computing platform to the location point filtered outIt utilizes
Commutative encryption algorithm encrypts again, obtains the ciphertext of double layer encryptionWherein I'NIndicate position in set H
Set title code name set a little, I'LIndicate the position coordinates set of location point in set H,The service of expression provides
Person is with commutative encryption public key to location point attribute (I'N,I'L) encryption ciphertext,It indicates to be used with recommendation
Family commutative encryption public key pairThe ciphertext of encryption.
9. according to the method described in claim 1, wherein recommended user calculates the predictive information received in step (8),
Refer to that recommended user carries out reversal to predictive information R', obtains its corresponding plaintext prediction and evaluation information Rp=Γ-1(R',s',
β), wherein s ' is a random number, and β is a Big prime.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510979917.5A CN105592085B (en) | 2015-12-23 | 2015-12-23 | The method for secret protection of facing position perception recommender system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510979917.5A CN105592085B (en) | 2015-12-23 | 2015-12-23 | The method for secret protection of facing position perception recommender system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105592085A CN105592085A (en) | 2016-05-18 |
| CN105592085B true CN105592085B (en) | 2018-12-14 |
Family
ID=55931299
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510979917.5A Active CN105592085B (en) | 2015-12-23 | 2015-12-23 | The method for secret protection of facing position perception recommender system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105592085B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106650485B (en) * | 2016-09-18 | 2019-06-28 | 山东大学 | Personalized environment based on Android perceives method for secret protection |
| CN109195098B (en) * | 2018-06-20 | 2020-11-03 | 苏州大学 | Shared collaborative filtering method based on differential privacy |
| CN109300540B (en) * | 2018-10-23 | 2021-10-15 | 北京理工大学 | Privacy protection medical service recommendation method in electronic medical system |
| CN109729077B (en) * | 2018-12-20 | 2021-03-23 | 西安电子科技大学 | Privacy protection method based on dynamic position association |
| CN109992995B (en) * | 2019-03-05 | 2021-05-14 | 华南理工大学 | Searchable encryption method supporting location protection and privacy inquiry |
| CN110149199B (en) * | 2019-05-22 | 2022-03-04 | 南京信息职业技术学院 | Privacy protection method and system based on attribute perception |
| CN110825955B (en) * | 2019-06-27 | 2024-06-25 | 安徽师范大学 | Distributed differential privacy recommendation method based on location service |
| CN113051587B (en) * | 2021-03-10 | 2024-02-02 | 中国人民大学 | Privacy protection intelligent transaction recommendation method, system and readable medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103826237A (en) * | 2014-02-28 | 2014-05-28 | 西安电子科技大学 | Method for establishing location privacy protection model for continuous location based service |
| US9094378B1 (en) * | 2013-08-16 | 2015-07-28 | Google Inc. | Homomorphic cryptography on numerical values in digital computing |
| CN105052070A (en) * | 2013-03-15 | 2015-11-11 | 三菱电机株式会社 | Method for authenticating encryption and system for authenticating biometric data |
-
2015
- 2015-12-23 CN CN201510979917.5A patent/CN105592085B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105052070A (en) * | 2013-03-15 | 2015-11-11 | 三菱电机株式会社 | Method for authenticating encryption and system for authenticating biometric data |
| US9094378B1 (en) * | 2013-08-16 | 2015-07-28 | Google Inc. | Homomorphic cryptography on numerical values in digital computing |
| CN103826237A (en) * | 2014-02-28 | 2014-05-28 | 西安电子科技大学 | Method for establishing location privacy protection model for continuous location based service |
Non-Patent Citations (1)
| Title |
|---|
| LBS中面向协同位置隐私保护的群组最近邻查询;高胜 等;《通信学报》;20150325;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105592085A (en) | 2016-05-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105592085B (en) | The method for secret protection of facing position perception recommender system | |
| US10043035B2 (en) | Systems and methods for enhancing data protection by anonosizing structured and unstructured data and incorporating machine learning and artificial intelligence in classical and quantum computing environments | |
| US20220012364A1 (en) | Systems and methods for enforcing privacy-respectful, trusted communications | |
| Wang et al. | A privacy-enhanced retrieval technology for the cloud-assisted internet of things | |
| Zhu et al. | Similarity-maintaining privacy preservation and location-aware low-rank matrix factorization for QoS prediction based web service recommendation | |
| Sun et al. | Data security and privacy in cloud computing | |
| CA2929269C (en) | Dynamic de-identification and anonymity | |
| US9087216B2 (en) | Dynamic de-identification and anonymity | |
| Ma et al. | APPLET: A privacy-preserving framework for location-aware recommender system | |
| CN113362048B (en) | Data label distribution determining method and device, computer equipment and storage medium | |
| Wang et al. | POI recommendation with federated learning and privacy preserving in cross domain recommendation | |
| US20220311754A1 (en) | Generating bridge match identifiers for linking identifers from server logs | |
| Liu et al. | A privacy-preserving framework for trust-oriented point-of-interest recommendation | |
| WO2019246568A1 (en) | Systems and methods for enforcing privacy-respectful, trusted communications | |
| Guo et al. | Enabling secure cross-modal retrieval over encrypted heterogeneous IoT databases with collective matrix factorization | |
| CN111490995A (en) | Model training method and device for protecting privacy, data processing method and server | |
| Kaleli et al. | SOM-based recommendations with privacy on multi-party vertically distributed data | |
| Kumar et al. | A hybrid solution for secure privacy-preserving cloud storage & information retrieval | |
| Yang et al. | Secure and privacy-preserving human interaction recognition of pervasive healthcare monitoring | |
| Nkenyereye et al. | Fog based secure framework for personal health records systems | |
| JP2017111793A (en) | Retrieval method, retrieval system and program | |
| KR102171458B1 (en) | Method of providing personal information collection agreement procedure in iot system, and apparatuses performing the same | |
| De Montjoye | Computational privacy: towards privacy-conscientious uses of metadata | |
| Amma et al. | P 2 CADL Privacy Preserving in Cloud using Autoencoder based Deep Learning Classifier for Smart City Applications | |
| JP5498974B2 (en) | Evaluation value estimation system and evaluation value estimation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |