CN113674036A - Recommendation method and system for mobile banking advertisements - Google Patents
Recommendation method and system for mobile banking advertisements Download PDFInfo
- Publication number
- CN113674036A CN113674036A CN202111039365.1A CN202111039365A CN113674036A CN 113674036 A CN113674036 A CN 113674036A CN 202111039365 A CN202111039365 A CN 202111039365A CN 113674036 A CN113674036 A CN 113674036A
- Authority
- CN
- China
- Prior art keywords
- advertisement
- scoring matrix
- advertisements
- matrix
- user
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 239000011159 matrix material Substances 0.000 claims abstract description 206
- 238000012549 training Methods 0.000 claims abstract description 43
- 238000012545 processing Methods 0.000 claims description 23
- 238000000354 decomposition reaction Methods 0.000 claims description 20
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 8
- 238000012804 iterative process Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0263—Targeted advertisements based upon Internet or website rating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q20/00—Payment architectures, schemes or protocols
- G06Q20/08—Payment architectures
- G06Q20/10—Payment architectures specially adapted for electronic funds transfer [EFT] systems; specially adapted for home banking systems
- G06Q20/108—Remote banking, e.g. home banking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0277—Online advertisement
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Telephone Function (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method and a system for recommending mobile banking advertisements, which can be applied to the field of big data. Acquiring the browsing times of a user for browsing each mobile phone bank advertisement; constructing a first advertisement scoring matrix according to the browsing times of the user for browsing the mobile phone bank advertisements; adding interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix; performing iterative training by using a preset algorithm based on the second advertisement scoring matrix until an iteration stopping condition is met, and obtaining the predicted browsing times of the mobile phone bank advertisements; recommending the top n mobile banking advertisements with the highest predicted browsing times to the user. The data of the user is protected, the mobile phone bank advertisements can be recommended accurately, and the safety of the user data and the user experience are improved.
Description
Technical Field
The invention relates to the technical field of big data, in particular to a method and a system for recommending mobile banking advertisements.
Background
At present, when the mobile banking pushes the mobile banking advertisement, the traditional recommendation algorithm is adopted, accurate pushing is not carried out according to the actual situation of the user, pushed advertisement information is uniform, and user experience is poor. In addition, non-banking applications introduced by mobile banking at present can also use related data of users, which may cause leakage of user data and poor security of user data.
Disclosure of Invention
In view of this, embodiments of the present invention provide a method and a system for recommending mobile banking advertisements, so as to solve the problems of poor user experience, poor security of user data, and the like in the existing manner of recommending mobile banking advertisements.
In order to achieve the above purpose, the embodiments of the present invention provide the following technical solutions:
the first aspect of the embodiment of the invention discloses a recommendation method for mobile banking advertisements, which comprises the following steps:
acquiring the browsing times of a user for browsing each mobile phone bank advertisement;
constructing a first advertisement scoring matrix according to the browsing times of the user for browsing each mobile phone bank advertisement, wherein each element in the first advertisement scoring matrix is the browsing times corresponding to the mobile phone bank advertisement;
adding interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix;
performing iterative training by using a preset algorithm based on the second advertisement scoring matrix until an iteration stopping condition is met, and obtaining the predicted browsing times of the mobile phone bank advertisements;
recommending the top n mobile banking advertisements with the highest predicted browsing times to the user.
Preferably, the adding interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix includes:
executing the following steps, processing each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix;
the following steps include:
for each element in the first advertisement scoring matrix, if the browsing times corresponding to the element are 0, the element is not processed, and if the browsing times corresponding to the element are not 0, laplacian noise is added to the element.
Preferably, the performing iterative training by using a preset algorithm based on the second advertisement scoring matrix until an iteration stop condition is met to obtain the predicted browsing times of each mobile banking advertisement includes:
adding the second advertisement scoring matrix into a matrix decomposition algorithm for iterative training until an iteration stop condition corresponding to a gradient descent algorithm is met, and obtaining a user data matrix and an advertisement data matrix;
and multiplying the user data matrix and the advertisement data matrix to obtain a third advertisement scoring matrix, wherein each element of the third advertisement scoring matrix is the predicted browsing times of the mobile banking advertisement.
Preferably, the method further comprises the following steps:
and in the process of adding the second advertisement scoring matrix into a matrix decomposition algorithm for iterative training, adding Laplace noise to an intermediate result obtained by each iteration.
Preferably, before recommending the top n number of the mobile banking advertisements with the highest predicted browsing times to the user, the method further includes:
and sequencing the mobile phone bank advertisements based on the predicted browsing times.
The second aspect of the embodiment of the invention discloses a recommendation system for mobile banking advertisements, which comprises:
the acquisition unit is used for acquiring the browsing times of the user for browsing the mobile phone bank advertisements;
the building unit is used for building a first advertisement scoring matrix according to the browsing times of the user for browsing the mobile phone bank advertisements, wherein each element in the first advertisement scoring matrix is the browsing times corresponding to the mobile phone bank advertisements;
the adding unit is used for adding interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix;
the processing unit is used for carrying out iterative training by using a preset algorithm based on the second advertisement scoring matrix until an iteration stopping condition is met, and obtaining the predicted browsing times of the mobile phone bank advertisements;
and the recommending unit is used for recommending the top n mobile banking advertisements with the highest predicted browsing times to the user.
Preferably, the adding unit is specifically configured to: executing the following steps, processing each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix;
the following steps include:
for each element in the first advertisement scoring matrix, if the browsing times corresponding to the element are 0, the element is not processed, and if the browsing times corresponding to the element are not 0, laplacian noise is added to the element.
Preferably, the processing unit includes:
the training module is used for adding the second advertisement scoring matrix into a matrix decomposition algorithm to carry out iterative training until an iteration stop condition corresponding to a gradient descent algorithm is met, and a user data matrix and an advertisement data matrix are obtained;
and the processing module is used for multiplying the user data matrix and the advertisement data matrix to obtain a third advertisement scoring matrix, and each element of the third advertisement scoring matrix is the predicted browsing times of the mobile banking advertisement.
Preferably, the processing unit further includes:
and the adding module is used for adding Laplace noise to the intermediate result obtained by each iteration in the process of adding the second advertisement scoring matrix to the matrix decomposition algorithm for iterative training.
Preferably, the system further comprises:
and the sorting unit is used for sorting the mobile phone bank advertisements based on the predicted browsing times.
Based on the method and the system for recommending mobile banking advertisements provided by the embodiment of the invention, the browsing times of the user for browsing the mobile banking advertisements are obtained; constructing a first advertisement scoring matrix according to the browsing times of the user for browsing the mobile phone bank advertisements; adding interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix; performing iterative training by using a preset algorithm based on the second advertisement scoring matrix until an iteration stopping condition is met, and obtaining the predicted browsing times of the mobile phone bank advertisements; recommending the top n mobile banking advertisements with the highest predicted browsing times to the user. According to the scheme, after an advertisement scoring matrix is built by using data of a user browsing mobile phone bank advertisements, interference noise is added to the built advertisement scoring matrix, then iterative training is carried out by using the disturbed advertisement scoring matrix to obtain the predicted browsing times of the mobile phone bank advertisements, and finally the mobile phone bank advertisements are recommended to the user according to the predicted browsing times, so that the mobile phone bank advertisements can be accurately recommended according to the data of the user, the safety of the user data is improved, and the user experience is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a flowchart of a method for recommending mobile banking advertisements according to an embodiment of the present invention;
FIG. 2 is a flowchart of obtaining predicted browsing times of mobile banking advertisements according to an embodiment of the present invention;
fig. 3 is a block diagram of a recommendation system for mobile banking advertisements according to an embodiment of the present invention;
fig. 4 is another block diagram of a recommendation system for mobile banking advertisements according to an embodiment of the present invention;
fig. 5 is a block diagram of another structure of a recommendation system for mobile banking advertisements according to an embodiment of the present invention;
fig. 6 is a block diagram of another structure of a recommendation system for mobile banking advertisements according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In this application, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
It should be noted that the method and the system for recommending mobile banking advertisements provided by the invention can be used in the field of big data. The above is only an example, and the application fields of the method and the system for recommending mobile banking advertisements provided by the present invention are not limited.
As can be seen from the background art, when the mobile banking pushes the mobile banking advertisement, the mobile banking does not push the advertisement accurately according to the actual situation of the user, and the pushed advertisement information is uniform, which results in poor user experience. In addition, non-banking applications introduced by mobile banking at present can also use related data of users, which may cause leakage of user data and poor security of user data.
Therefore, the embodiment of the invention provides a method and a system for recommending mobile banking advertisements, wherein after an advertisement scoring matrix is constructed by using data of mobile banking advertisements browsed by a user, interference noise is added to the constructed advertisement scoring matrix, then iterative training is performed by using the disturbed advertisement scoring matrix to obtain the predicted browsing times of the mobile banking advertisements, and finally the mobile banking advertisements are recommended to the user according to the predicted browsing times, so that the mobile banking advertisements can be accurately recommended by protecting the data of the user, and the safety of the user data and the user experience are improved.
Referring to fig. 1, a flowchart of a recommendation method for mobile banking advertisements provided by an embodiment of the present invention is shown, where the recommendation method includes:
step S101: and acquiring the browsing times of the user for browsing the mobile phone bank advertisements.
It should be noted that, the mobile banking institute may provide a plurality of mobile banking advertisements to the user, and a user clicking a certain mobile banking advertisement may be regarded as the user browsing the mobile banking advertisement.
In the process of implementing step S101 specifically, browsing history data of the user is acquired, and according to the acquired browsing history data, browsing times of the user browsing each mobile banking advertisement can be determined.
It is understood that, for a certain mobile banking advertisement provided to a user, the browsing times of the user browsing the mobile banking advertisement may be 0 times (i.e. the user does not click to browse the mobile banking advertisement), and the browsing times of the user browsing the mobile banking advertisement may not be 0 times (i.e. the user clicks to browse the mobile banking advertisement).
That is, after the processing of step S101, the determined number of views corresponding to a certain mobile banking advertisement may or may not be 0.
Step S102: and constructing a first advertisement scoring matrix according to the browsing times of the user for browsing the mobile phone bank advertisements.
It should be noted that each element in the first advertisement scoring matrix is the browsing times corresponding to the mobile banking advertisement.
In the process of implementing step S102 specifically, a first advertisement scoring matrix is constructed by using the browsing times of the user browsing each mobile banking advertisement, and each element in the first advertisement scoring matrix corresponds to the browsing times corresponding to one mobile banking advertisement. That is, the first advertisement scoring matrix includes the browsing times corresponding to each mobile banking advertisement (where the browsing times corresponding to a certain mobile banking advertisement may or may not be 0), for example: assuming that the browsing times corresponding to the mobile banking advertisement A are 0 and the browsing times corresponding to the mobile banking advertisement B are 10, the value of the element corresponding to the mobile banking advertisement A in the first advertisement scoring matrix is 0, and the value of the element corresponding to the mobile banking advertisement B is 10.
It can be understood that, because the types and the number of the mobile banking advertisements provided to the user are large, the mobile banking advertisements interested by the user only relate to a small range, and therefore, the values of more elements in the first advertisement scoring matrix are 0, that is, more mobile banking advertisements are not viewed by the user, and the constructed first advertisement scoring matrix can be regarded as a sparse matrix.
It can be understood that, in the first advertisement scoring matrix, rows identify users, columns identify mobile banking advertisements, and assuming that m users and m mobile banking advertisements exist, the first advertisement scoring matrix is an m × m matrix, and since a certain user only browses certain mobile banking advertisements in the m mobile banking advertisements (that is, the mobile banking advertisements in which the user is interested only relate to a small range), the first advertisement scoring matrix can be regarded as a sparse matrix.
Step S103: and adding interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix.
In the process of specifically implementing step S103, an interference noise is added to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix, and a specific manner of adding an interference noise to each element in the first advertisement scoring matrix is as follows: for each element in the first advertisement scoring matrix, if the browsing frequency corresponding to the element is 0, the element is not processed, and if the browsing frequency corresponding to the element is not 0, laplacian noise is added to the element.
That is, no interference noise is added to the elements having a value of 0 in the first advertisement scoring matrix, and laplacian noise is added to the elements having a value other than 0 in the first advertisement scoring matrix. By the method, privacy protection can be performed on the data contained in the first advertisement scoring matrix, accurate data cannot be obtained even if other applications intercept the first advertisement scoring matrix, and the user data can be protected.
And after the elements in the first advertisement scoring matrix are processed, a second advertisement scoring matrix can be obtained.
In some embodiments, in the process of adding laplacian noise to the elements having values other than 0 in the first advertisement scoring matrix, the laplacian noise function (or laplacian method) may be specifically utilized to add laplacian noise to the elements having values other than 0. When the laplacian noise is added by using the laplacian noise function, the privacy parameter and the sensitivity parameter need to be determined according to the actual situation to prevent the added laplacian noise from being too large or too small, that is, the laplacian noise is added on the premise of ensuring the safety and the usability of data. Specifically, for a first advertisement scoring matrix, a laplacian noise matrix with the same dimension as the first advertisement scoring matrix can be generated through a laplacian noise function, and a second advertisement scoring matrix can be obtained by adding the first advertisement scoring matrix and the laplacian noise matrix. For example: and assuming that the first advertisement scoring matrix is an m matrix, generating an m Laplace noise matrix through a Laplace noise function, and adding the m Laplace noise matrix and the m Laplace noise matrix to obtain a second advertisement scoring matrix.
Step S104: and based on the second advertisement scoring matrix, performing iterative training by using a preset algorithm until an iteration stopping condition is met, and obtaining the predicted browsing times of the mobile phone bank advertisements.
In the process of implementing step S104, the second advertisement scoring matrix (i.e., the first advertisement scoring matrix added with the interference noise) is added to the preset algorithm to perform iterative training until the iteration stop condition is satisfied, so as to obtain the predicted browsing times of the advertisements of each mobile banking bank.
It can be understood from the foregoing content of step S102 that there are multiple elements with a value of 0 (i.e. the mobile banking advertisements with a browsing number of 0) in the first advertisement scoring matrix, and through the processing of step S104, the predicted browsing number of all the mobile banking advertisements (including the mobile banking advertisements with a previous browsing number of 0) can be obtained, and the predicted browsing number of a certain mobile banking advertisement can indicate the number of times that the mobile banking advertisement is likely to be browsed by the user.
In some embodiments, the preset algorithm for performing the iterative training may be a matrix decomposition algorithm, and the iteration stop condition may be a training termination condition (including at least the number of iterations and the amount of change) corresponding to the gradient descent algorithm.
Specifically, the second advertisement scoring matrix is used as training data, iterative training is performed through a matrix decomposition algorithm and a gradient descent algorithm, and finally the predicted browsing times of the advertisements of the mobile phone banks can be obtained.
Preferably, after obtaining the predicted browsing times of the mobile banking advertisements, the mobile banking advertisements are sorted based on the predicted browsing times, for example: the advertisements of each mobile phone bank can be sequenced according to the sequence of the predicted browsing times from high to low.
Step S105: recommending the top n mobile banking advertisements with the highest predicted browsing times to the user.
In the process of implementing step S105 specifically, the first n mobile banking advertisements with the highest predicted browsing times are recommended to the user, where n is an integer greater than 0.
It can be understood that after the mobile phone bank advertisements are sequenced according to the predicted browsing times, the top n mobile phone bank advertisements with the highest predicted browsing times are selected as the mobile phone bank advertisements which are most likely to be interested by the user, and the top n mobile phone bank advertisements with the highest predicted browsing times are recommended to the user, so that the purpose of accurate recommendation is achieved.
In the embodiment of the invention, after the advertisement scoring matrix is constructed by using the data of the mobile phone bank advertisement browsed by the user, the interference noise is added to the constructed advertisement scoring matrix, the disturbed advertisement scoring matrix is used for iterative training to obtain the predicted browsing times of the mobile phone bank advertisement, and finally the mobile phone bank advertisement is recommended to the user according to the predicted browsing times, so that the mobile phone bank advertisement can be accurately recommended by using the data of the user, the safety of the user data is improved, and the user experience is improved.
In the above embodiment of the present invention, referring to fig. 2, the process of obtaining the predicted browsing times of each mobile banking advertisement in step S104 in fig. 1 shows a flowchart of obtaining the predicted browsing times of the mobile banking advertisement provided in the embodiment of the present invention, which includes the following steps:
step S201: and adding the second advertisement scoring matrix into the matrix decomposition algorithm for iterative training until an iteration stop condition corresponding to the gradient descent algorithm is met, and obtaining a user data matrix and an advertisement data matrix.
In the process of specifically implementing step S201, a second advertisement scoring matrix (that is, the first advertisement scoring matrix added with the interference noise) is added to the matrix decomposition algorithm for iterative training until the iterative training satisfies an iteration stop condition corresponding to the gradient descent algorithm, where a final result obtained at this time includes two matrices, each of which is: a user data matrix and an advertisement data matrix.
It should be noted that the main function of the matrix decomposition algorithm is a Loss function (such as a flat method).
It can be understood that, in the process of adding the second advertisement scoring matrix to the matrix decomposition algorithm for iterative training, multiple iterations are required, each iteration will obtain a corresponding intermediate result (non-final result), and in order to further ensure the security of the user data, it is preferable that, in the process of adding the second advertisement scoring matrix to the matrix decomposition algorithm for iterative training, laplacian noise is added to the intermediate result obtained by each iteration. By the method, the safety of the user data is further improved.
In some embodiments, in the process of adding the second advertisement scoring matrix to the matrix decomposition algorithm for iterative training, each iteration obtains an intermediate result of the user data matrix and the advertisement data matrix corresponding to the iteration, at this time, a laplacian noise matrix with the same dimension as the intermediate result is synchronously generated, a prediction error value of the current iteration is calculated according to the intermediate result, and the prediction error value is introduced into the gradient descent algorithm and used for calculating the user data matrix and the advertisement data matrix of the next iteration until the iteration is finished.
Step S202: and multiplying the user data matrix and the advertisement data matrix to obtain a third advertisement scoring matrix.
It should be noted that each element of the third advertisement scoring matrix is the predicted browsing times of the mobile banking advertisement.
In step S201, the final result obtained by the iterative training includes two matrices, i.e., a user data matrix and an advertisement data matrix, and in the process of implementing step S202, the user data matrix and the advertisement data matrix are multiplied to obtain a third advertisement scoring matrix, where each element in the third advertisement scoring matrix is the predicted browsing times of the mobile banking advertisement.
That is, the third advertisement scoring matrix obtained by multiplying the user data matrix and the advertisement data matrix includes the predicted browsing times of the mobile banking advertisements.
As can be seen from the content in fig. 1 in the embodiment of the present invention, each row element in the third advertisement scoring matrix represents the predicted browsing times of a certain user for all mobile banking advertisements, and the first n mobile banking advertisements with the highest predicted browsing times in the row element can be selected and recommended to the user corresponding to the row element.
In the embodiment of the invention, the second advertisement scoring matrix obtained by adding interference noise processing is added into a matrix decomposition algorithm for iterative training, Laplace noise is added to the intermediate result obtained by each iteration in the iterative process until the iteration stop condition corresponding to the gradient descent algorithm is met, a third advertisement scoring matrix containing the predicted browsing times of each mobile phone bank advertisement is obtained, and finally the mobile phone bank advertisement is recommended to the user according to the predicted browsing times, so that the data of the user is further protected, the mobile phone bank advertisement can be accurately recommended, the safety of the user data is improved, and the user experience is improved.
Corresponding to the method for recommending mobile banking advertisements provided in the embodiment of the present invention, referring to fig. 3, an embodiment of the present invention further provides a structural block diagram of a system for recommending mobile banking advertisements, where the system for recommending mobile banking advertisements includes: an acquisition unit 301, a construction unit 302, an addition unit 303, a processing unit 304, and a recommendation unit 305;
the obtaining unit 301 is configured to obtain browsing times of browsing each mobile banking advertisement by a user.
The constructing unit 302 is configured to construct a first advertisement scoring matrix according to the browsing times of the user browsing each mobile banking advertisement, where each element in the first advertisement scoring matrix is the browsing time corresponding to the mobile banking advertisement.
The adding unit 303 is configured to add interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix.
In a specific implementation, the adding unit 303 is specifically configured to: executing the following steps, processing each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix; the method comprises the following steps: for each element in the first advertisement scoring matrix, if the browsing times corresponding to the element are 0, the element is not processed, and if the browsing times corresponding to the element are not 0, laplacian noise is added to the element.
And the processing unit 304 is configured to perform iterative training by using a preset algorithm based on the second advertisement scoring matrix until an iteration stop condition is met, so as to obtain the predicted browsing times of the mobile banking advertisements.
And the recommending unit 305 is configured to recommend the top n mobile banking advertisements with the highest predicted browsing times to the user.
In the embodiment of the invention, after the advertisement scoring matrix is constructed by using the data of the mobile phone bank advertisement browsed by the user, the interference noise is added to the constructed advertisement scoring matrix, the disturbed advertisement scoring matrix is used for iterative training to obtain the predicted browsing times of the mobile phone bank advertisement, and finally the mobile phone bank advertisement is recommended to the user according to the predicted browsing times, so that the mobile phone bank advertisement can be accurately recommended by using the data of the user, the safety of the user data is improved, and the user experience is improved.
Preferably, referring to fig. 4 in conjunction with fig. 3, another structural block diagram of a recommendation system for mobile banking advertisements provided in an embodiment of the present invention is shown, where the processing unit 304 includes: a training module 3041 and a processing module 3042;
the training module 3041 is configured to add the second advertisement scoring matrix to the matrix decomposition algorithm for iterative training until an iteration stop condition corresponding to the gradient descent algorithm is met, so as to obtain a user data matrix and an advertisement data matrix.
The processing module 3042 is configured to multiply the user data matrix and the third advertisement scoring matrix to obtain a third advertisement scoring matrix, where each element of the third advertisement scoring matrix is a predicted browsing frequency of the mobile banking advertisement.
Preferably, referring to fig. 5 in conjunction with fig. 4, a further structural block diagram of a recommendation system for mobile banking advertisements provided in an embodiment of the present invention is shown, where the processing unit 304 further includes: an add module 3043;
the adding module 3043 is configured to add laplacian noise to an intermediate result obtained by each iteration in the process of adding the second advertisement scoring matrix to the matrix decomposition algorithm for iterative training.
In the embodiment of the invention, the second advertisement scoring matrix obtained by adding interference noise processing is added into a matrix decomposition algorithm for iterative training, Laplace noise is added to the intermediate result obtained by each iteration in the iterative process until the iteration stop condition corresponding to the gradient descent algorithm is met, a third advertisement scoring matrix containing the predicted browsing times of each mobile phone bank advertisement is obtained, and finally the mobile phone bank advertisement is recommended to the user according to the predicted browsing times, so that the data of the user is further protected, the mobile phone bank advertisement can be accurately recommended, the safety of the user data is improved, and the user experience is improved.
Preferably, referring to fig. 6 in conjunction with fig. 3, there is shown another structural block diagram of a recommendation system for mobile banking advertisements according to an embodiment of the present invention, where the recommendation system further includes:
and the sorting unit 306 is used for sorting the mobile banking advertisements based on the predicted browsing times.
In summary, embodiments of the present invention provide a method and a system for recommending mobile banking advertisements, where after an advertisement scoring matrix is constructed by using data of a user browsing mobile banking advertisements, interference noise is added to the constructed advertisement scoring matrix, then iterative training is performed by using the disturbed advertisement scoring matrix to obtain predicted browsing times of the mobile banking advertisements, and finally the mobile banking advertisements are recommended to the user according to the predicted browsing times, so that the mobile banking advertisements can be accurately recommended while protecting the data of the user, and the security of the user data and the user experience are improved.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, the system or system embodiments are substantially similar to the method embodiments and therefore are described in a relatively simple manner, and reference may be made to some of the descriptions of the method embodiments for related points. The above-described system and system embodiments are only illustrative, wherein the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A recommendation method for mobile banking advertisements is characterized by comprising the following steps:
acquiring the browsing times of a user for browsing each mobile phone bank advertisement;
constructing a first advertisement scoring matrix according to the browsing times of the user for browsing each mobile phone bank advertisement, wherein each element in the first advertisement scoring matrix is the browsing times corresponding to the mobile phone bank advertisement;
adding interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix;
performing iterative training by using a preset algorithm based on the second advertisement scoring matrix until an iteration stopping condition is met, and obtaining the predicted browsing times of the mobile phone bank advertisements;
recommending the top n mobile banking advertisements with the highest predicted browsing times to the user.
2. The method of claim 1, wherein adding interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix comprises:
executing the following steps, processing each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix;
the following steps include:
for each element in the first advertisement scoring matrix, if the browsing times corresponding to the element are 0, the element is not processed, and if the browsing times corresponding to the element are not 0, laplacian noise is added to the element.
3. The method according to claim 1, wherein the obtaining of the predicted browsing times of each mobile banking advertisement by performing iterative training by using a preset algorithm until an iteration stop condition is met based on the second advertisement scoring matrix comprises:
adding the second advertisement scoring matrix into a matrix decomposition algorithm for iterative training until an iteration stop condition corresponding to a gradient descent algorithm is met, and obtaining a user data matrix and an advertisement data matrix;
and multiplying the user data matrix and the advertisement data matrix to obtain a third advertisement scoring matrix, wherein each element of the third advertisement scoring matrix is the predicted browsing times of the mobile banking advertisement.
4. The method of claim 3, further comprising:
and in the process of adding the second advertisement scoring matrix into a matrix decomposition algorithm for iterative training, adding Laplace noise to an intermediate result obtained by each iteration.
5. The method of claim 1, wherein before recommending the top n number of the cell phone bank advertisements with the highest number of predicted views to the user, further comprising:
and sequencing the mobile phone bank advertisements based on the predicted browsing times.
6. A recommendation system for mobile banking advertisements is characterized by comprising:
the acquisition unit is used for acquiring the browsing times of the user for browsing the mobile phone bank advertisements;
the building unit is used for building a first advertisement scoring matrix according to the browsing times of the user for browsing the mobile phone bank advertisements, wherein each element in the first advertisement scoring matrix is the browsing times corresponding to the mobile phone bank advertisements;
the adding unit is used for adding interference noise to each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix;
the processing unit is used for carrying out iterative training by using a preset algorithm based on the second advertisement scoring matrix until an iteration stopping condition is met, and obtaining the predicted browsing times of the mobile phone bank advertisements;
and the recommending unit is used for recommending the top n mobile banking advertisements with the highest predicted browsing times to the user.
7. The system according to claim 6, wherein the adding unit is specifically configured to: executing the following steps, processing each element in the first advertisement scoring matrix to obtain a second advertisement scoring matrix;
the following steps include:
for each element in the first advertisement scoring matrix, if the browsing times corresponding to the element are 0, the element is not processed, and if the browsing times corresponding to the element are not 0, laplacian noise is added to the element.
8. The system of claim 6, wherein the processing unit comprises:
the training module is used for adding the second advertisement scoring matrix into a matrix decomposition algorithm to carry out iterative training until an iteration stop condition corresponding to a gradient descent algorithm is met, and a user data matrix and an advertisement data matrix are obtained;
and the processing module is used for multiplying the user data matrix and the advertisement data matrix to obtain a third advertisement scoring matrix, and each element of the third advertisement scoring matrix is the predicted browsing times of the mobile banking advertisement.
9. The system of claim 8, wherein the processing unit further comprises:
and the adding module is used for adding Laplace noise to the intermediate result obtained by each iteration in the process of adding the second advertisement scoring matrix to the matrix decomposition algorithm for iterative training.
10. The system of claim 6, further comprising:
and the sorting unit is used for sorting the mobile phone bank advertisements based on the predicted browsing times.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111039365.1A CN113674036A (en) | 2021-09-06 | 2021-09-06 | Recommendation method and system for mobile banking advertisements |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111039365.1A CN113674036A (en) | 2021-09-06 | 2021-09-06 | Recommendation method and system for mobile banking advertisements |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113674036A true CN113674036A (en) | 2021-11-19 |
Family
ID=78548753
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111039365.1A Pending CN113674036A (en) | 2021-09-06 | 2021-09-06 | Recommendation method and system for mobile banking advertisements |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113674036A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107392049A (en) * | 2017-07-26 | 2017-11-24 | 安徽大学 | Recommendation method based on differential privacy protection |
| WO2019056573A1 (en) * | 2017-09-25 | 2019-03-28 | 深圳大学 | Differential privacy-based system and method for collaborative web quality-of-service prediction for privacy protection |
| WO2019056572A1 (en) * | 2017-09-25 | 2019-03-28 | 深圳大学 | Model-based collaborative filtering method for collaborative web quality-of-service prediction for privacy protection |
| CN110837603A (en) * | 2019-11-09 | 2020-02-25 | 安徽大学 | Integrated recommendation method based on differential privacy protection |
-
2021
- 2021-09-06 CN CN202111039365.1A patent/CN113674036A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107392049A (en) * | 2017-07-26 | 2017-11-24 | 安徽大学 | Recommendation method based on differential privacy protection |
| WO2019056573A1 (en) * | 2017-09-25 | 2019-03-28 | 深圳大学 | Differential privacy-based system and method for collaborative web quality-of-service prediction for privacy protection |
| WO2019056572A1 (en) * | 2017-09-25 | 2019-03-28 | 深圳大学 | Model-based collaborative filtering method for collaborative web quality-of-service prediction for privacy protection |
| CN110837603A (en) * | 2019-11-09 | 2020-02-25 | 安徽大学 | Integrated recommendation method based on differential privacy protection |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Goh et al. | Classification of scale-free networks | |
| Chung | The heat kernel as the pagerank of a graph | |
| US10402479B2 (en) | Method, server, browser, and system for recommending text information | |
| CN103763361A (en) | Method and system for recommending applications based on user behavior and recommending server | |
| WO2014194689A1 (en) | Method, server, browser, and system for recommending text information | |
| CN105005582A (en) | Recommendation method and device for multimedia information | |
| Jia et al. | Evidence of a decreasing trend for the Hubble constant | |
| CN108446297B (en) | Recommendation method and device and electronic equipment | |
| CN104376083A (en) | Graph recommendation method based on concern relations and multiple user behaviors | |
| Yan et al. | Spectral techniques and mathematical aspects of K 4 chain graph | |
| CN109977316A (en) | A kind of parallel type article recommended method, device, equipment and storage medium | |
| CN110020124A (en) | The method and device excavated for Related product | |
| CN105095258B (en) | A kind of media information sort method, device and media information recommending system | |
| CN111797319A (en) | Recommendation method, device, equipment and storage medium | |
| CN107798558A (en) | User behavior data analysis method and server | |
| CN107369066B (en) | Feature comparison method and device between comment objects | |
| Smith et al. | Chromospheric magnetic reconnection: two-fluid simulations of coalescing current loops | |
| CN111369324B (en) | Target information determining method, device, equipment and readable storage medium | |
| CN113674036A (en) | Recommendation method and system for mobile banking advertisements | |
| CN104199843A (en) | Service sorting and recommending method and system based on social network interactive data | |
| CN111666481B (en) | Data mining method, device, computer readable medium and electronic equipment | |
| CN108763515B (en) | Time-sensitive personalized recommendation method based on probability matrix decomposition | |
| CN111353001B (en) | Method and device for classifying users | |
| CN116485503A (en) | Commodity combination recommendation method, device, equipment and medium thereof | |
| CN113674066A (en) | Recommendation method and system for mobile banking financing products |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |