CN112700281A - Behavior prediction method, behavior prediction device, behavior prediction equipment and computer readable storage medium - Google Patents
Behavior prediction method, behavior prediction device, behavior prediction equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN112700281A CN112700281A CN202011644642.7A CN202011644642A CN112700281A CN 112700281 A CN112700281 A CN 112700281A CN 202011644642 A CN202011644642 A CN 202011644642A CN 112700281 A CN112700281 A CN 112700281A
- Authority
- CN
- China
- Prior art keywords
- user
- behavior prediction
- predicted
- feature
- prediction
- 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
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
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/60—Protecting data
- G06F21/602—Providing cryptographic facilities or services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Software Systems (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Bioethics (AREA)
- General Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Computer Security & Cryptography (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application provides a behavior prediction method, a behavior prediction device, equipment and a computer-readable storage medium, which are applied to a first participant of federal learning, wherein the method comprises the following steps: inputting the obtained sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model; receiving a behavior prediction request sent by a user terminal; acquiring user characteristics of a user to be predicted based on the behavior prediction request; inputting the user characteristics into a trained behavior prediction model to obtain at least one prediction result, wherein the prediction result comprises the probability that the user to be predicted goes to a destination to be predicted; and sending the at least one prediction result to the user terminal so that the user terminal outputs the at least one prediction result to realize the prediction of the behavior intention of the user to be predicted, and when the prediction result is applied to a marketing scene, a more appropriate store can be recommended to the user in a targeted manner, the purchasing experience of the user is improved, and the sales volume is increased.
Description
Technical Field
The present application relates to the field of machine learning technology, and relates to, but is not limited to, a behavior prediction method, apparatus, device, and computer-readable storage medium.
Background
With the rapid development of internet technology, more and more information is acquired by browsing the internet. In a marketing scenario, in a scheme for predicting customer-to-store in the related art, a separate model needs to be trained for each store, a server needs to predict whether a customer arrives at a store once according to the model of each store to obtain a plurality of prediction results, and one store is selected to be allocated to a customer with a purchase intention by combining the plurality of prediction results. The customer behavior prediction mode needs to manage a large number of models, and brings inconvenience to store management and server prediction; moreover, a large number of models need to be called during prediction, so that serious burden is brought to a system; meanwhile, a plurality of prediction results are output by a plurality of models respectively, and further integration and comparison are needed to obtain a target store, so that the prediction efficiency is low.
Disclosure of Invention
Embodiments of the present application provide a behavior prediction method, apparatus, device, computer-readable storage medium, and computer program product, which can quickly and accurately predict a target store for a clue customer.
The technical scheme of the embodiment of the application is realized as follows:
the embodiment of the application provides a behavior prediction method, which is applied to a first participant of federal learning, and comprises the following steps:
acquiring a sample feature set, inputting the sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model, wherein the behavior prediction model is a model based on federal learning training;
receiving a behavior prediction request sent by a user terminal;
acquiring user characteristics of a user to be predicted based on the behavior prediction request;
inputting the user characteristics into a trained behavior prediction model to obtain at least one prediction result, wherein the prediction result comprises the probability that the user to be predicted goes to a destination to be predicted;
and sending the at least one prediction result to the user terminal so that the user terminal outputs the at least one prediction result.
The embodiment of the application provides a behavior prediction device, which is applied to a first participant of federal learning, and the device comprises:
the training module is used for acquiring a sample feature set, inputting the sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model, and the behavior prediction model is a model based on federal learning training;
the receiving module is used for receiving a behavior prediction request sent by a user terminal;
the first obtaining module is used for obtaining the user characteristics of the user to be predicted based on the behavior prediction request;
the input module is used for inputting the user characteristics into a trained behavior prediction model to obtain at least one prediction result, and the prediction result comprises the probability that the user to be predicted goes to a destination to be predicted;
and the sending module is used for sending the at least one prediction result to the user terminal so as to enable the user terminal to output the at least one prediction result.
The embodiment of the application provides a behavior prediction device, which is applied to a first participant of federal learning, and comprises:
a memory for storing executable instructions;
and the processor is used for realizing the method provided by the embodiment of the application when executing the executable instructions stored in the memory.
Embodiments of the present application provide a computer-readable storage medium, where executable instructions are stored on the computer-readable storage medium, and when the computer-readable storage medium is executed by a processor, the computer-readable storage medium implements a method provided by embodiments of the present application.
Embodiments of the present application provide a computer program product, which includes a computer program, and when the computer program is executed by a processor, the computer program implements the method provided by the embodiments of the present application.
The embodiment of the application has the following beneficial effects:
in the behavior prediction method provided by the embodiment of the application, a sample feature set is obtained firstly, and the sample feature set is input into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model, wherein the behavior prediction model is a model based on federal learning training; after receiving a behavior prediction request sent by a user terminal; acquiring user characteristics of a user to be predicted based on the behavior prediction request; inputting the user characteristics into a trained behavior prediction model to obtain at least one prediction result, wherein the prediction result comprises the probability that the user to be predicted goes to a destination to be predicted; and finally, the at least one prediction result is sent to the user terminal, so that the user terminal outputs the at least one prediction result, the behavior intention of the user to be predicted is predicted, and the prediction method is applied to a marketing scene, can recommend a more appropriate store to the user in a targeted manner, improves the purchasing experience of the user and further improves the sales volume.
Drawings
Fig. 1 is a schematic diagram of a network architecture of a behavior prediction method according to an embodiment of the present application;
fig. 2 is a schematic structural diagram of a behavior prediction device provided in an embodiment of the present application;
fig. 3 is a schematic flow chart of an implementation of a behavior prediction method according to an embodiment of the present application;
fig. 4 is a schematic flow chart of another implementation of the behavior prediction method according to the embodiment of the present application;
fig. 5 is a schematic flow chart of another implementation of the behavior prediction method according to the embodiment of the present application;
FIG. 6 is a schematic diagram of a model training phase provided by an embodiment of the present application;
fig. 7 is a schematic diagram of a prediction phase provided in an embodiment of the present application.
Detailed Description
In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be described in further detail with reference to the attached drawings, the described embodiments should not be considered as limiting the present application, and all other embodiments obtained by a person of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.
In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict.
In the following description, references to the terms "first \ second \ third" are only used to distinguish similar objects and do not denote a particular order, but rather the terms "first \ second \ third" are used to interchange specific orders or sequences, where permissible, so that the embodiments of the present application described herein can be practiced in other than the order shown or described herein.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the present application only and is not intended to be limiting of the application.
Before further detailed description of the embodiments of the present application, terms and expressions referred to in the embodiments of the present application will be described, and the terms and expressions referred to in the embodiments of the present application will be used for the following explanation.
Before further detailed description of the embodiments of the present application, terms and expressions referred to in the embodiments of the present application will be described, and the terms and expressions referred to in the embodiments of the present application will be used for the following explanation.
1) Federal Learning (fed Learning), an emerging artificial intelligence base technology, is designed to carry out efficient machine Learning among multiple participants or multiple computing nodes on the premise of guaranteeing information security during big data exchange, protecting terminal data and personal data privacy and guaranteeing legal compliance.
2) In Vertical federal Learning (Vertical fed Learning), under the condition that the users of two data sets overlap more and the user features overlap less, the data sets are segmented according to the Vertical direction (namely feature dimension), and partial data which are the same for both users but have not the same user features are taken out for training machine Learning.
3) Homomorphic Encryption (Homomorphic Encryption), which is a cryptographic technique based on the theory of computational complexity of mathematical puzzles. The homomorphic encrypted data is processed to produce an output, which is decrypted, the result being the same as the output obtained by processing the unencrypted original data in the same way.
4) A thread client refers to a client who has purchasing intent in a marketing scenario.
An exemplary application of the apparatus implementing the embodiments of the present application, which may be implemented as a behavior prediction device, is described below. In the following, exemplary applications covered when the apparatus is implemented as a behavior prediction device will be explained.
Fig. 1 is a schematic diagram of a network architecture of a behavior prediction method according to an embodiment of the present application, and as shown in fig. 1, the network architecture at least includes a behavior prediction device 100 applied to a first party of federal learning, a network 200, a user terminal 300, and a server 400 applied to a second party of federal learning. In order to support an exemplary application, the behavior prediction device 100 may be a server, or may be a terminal device such as a desktop computer or a notebook computer, for performing behavior prediction and obtaining a prediction result. The user terminal 300 is a terminal that initiates a prediction request, and may be a mobile phone (mobile phone), a tablet computer, a notebook computer, or the like. The server 400 is a device for training the behavior prediction model, and may be a server, or a terminal device such as a desktop computer or a notebook computer. The behavior prediction device 100 connects the user terminal 300 and the server 400 through the network 200, and the network 200 may be a wide area network or a local area network, or a combination of the two, and uses a wireless or wired link to realize data transmission.
First, based on the federal learning training behavior prediction model, the behavior prediction apparatus 100 generates a public key and a private key for homomorphic encryption and decryption, and transmits the public key to the server 400. The server 400 encrypts the second feature of the server based on the public key to obtain a second ciphertext feature, returns the second ciphertext feature to the behavior prediction device 100, and the behavior prediction device 100 decrypts the second ciphertext feature according to the private key to obtain a second encoding value ZB. Meanwhile, the behavior prediction apparatus 100 determines the first coded value Z according to the first characteristic thereofAIs a reaction of ZAAnd ZBAnd sent to the server 400. Server 400 to ZBCoding to obtain a coded value DBObtaining DBAnd ZAAnd obtaining feature intersection. After the behavior prediction device 100 receives the feature intersection, it bases on the feature intersection and ZAAnd determining sample characteristics, and inputting the sample characteristics into a behavior prediction model constructed based on federal learning for training to obtain the trained behavior prediction model.
After the trained behavior prediction model is obtained, the user behavior can be predicted. The user terminal 300 sends a behavior prediction request to the behavior prediction device 100, the behavior prediction device 100 obtains user characteristics of a user to be predicted based on the received behavior prediction request, then inputs the user characteristics to a trained behavior prediction model to obtain a prediction result, the prediction result comprises the probability that the user to be predicted goes to a destination to be predicted, finally, the prediction result is sent to the user terminal 300, and the user terminal 300 outputs the prediction result. The behavior prediction method provided by the embodiment of the application can realize prediction of behavior intention of the user to be predicted, is applied to a marketing scene, can recommend a more appropriate store to the user in a targeted manner, and improves user purchasing experience, so that sales volume is increased.
The apparatus provided in the embodiments of the present application may be implemented as hardware or a combination of hardware and software, and various exemplary implementations of the apparatus provided in the embodiments of the present application are described below.
According to the exemplary structure of the behavior prediction device shown in fig. 2, the behavior prediction device is shown by taking the behavior prediction device 100 as an example, other exemplary structures of the behavior prediction device can be foreseen, and therefore, the structure described herein should not be considered as a limitation, for example, some components described below may be omitted, or components not described below may be added to adapt to the special requirements of some applications.
The behavior prediction apparatus 100 shown in fig. 2 includes: at least one processor 110, memory 140, at least one network interface 120, and a user interface 130. Each of the components in the behavior prediction device 100 are coupled together by a bus system 150. It will be appreciated that the bus system 150 is used to enable communications among the components of the connection. The bus system 150 includes a power bus, a control bus, and a status signal bus in addition to a data bus. For clarity of illustration, however, the various buses are labeled as bus system 150 in fig. 2.
The user interface 130 may include a display, a keyboard, a mouse, a touch-sensitive pad, a touch screen, and the like.
The memory 140 may be either volatile memory or nonvolatile memory, and may also include both volatile and nonvolatile memory. The nonvolatile Memory may be a Read Only Memory (ROM). The volatile Memory may be Random Access Memory (RAM). The memory 140 described in embodiments herein is intended to comprise any suitable type of memory.
The memory 140 in the embodiment of the present application is capable of storing data to support the operation of the behavior prediction apparatus 100. Examples of such data include: any computer program, such as an operating system and an application program, for operating on the behavior prediction device 100. The operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, and the like, and is used for implementing various basic services and processing hardware-based tasks. The application program may include various application programs.
As an example of the method provided by the embodiment of the present application implemented by software, the method provided by the embodiment of the present application may be directly embodied as a combination of software modules executed by the processor 110, the software modules may be located in a storage medium located in the memory 140, and the processor 110 reads executable instructions included in the software modules in the memory 140, and completes the method provided by the embodiment of the present application in combination with necessary hardware (for example, including the processor 110 and other components connected to the bus 150).
By way of example, the Processor 110 may be an integrated circuit chip having Signal processing capabilities, such as a general purpose Processor, a Digital Signal Processor (DSP), or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or the like, wherein the general purpose Processor may be a microprocessor or any conventional Processor or the like.
The behavior prediction method provided by the embodiment of the present application will be described in conjunction with exemplary applications and implementations of the behavior prediction device provided by the embodiment of the present application.
Fig. 3 is a schematic implementation flow diagram of a behavior prediction method according to an embodiment of the present application, which is applied to a behavior prediction device of the network architecture shown in fig. 1, and will be described with reference to the steps shown in fig. 3.
Step S301, a sample feature set is obtained and input into a pre-constructed behavior prediction model for training, so as to obtain a trained behavior prediction model.
Here, the behavior prediction model is a model trained based on federal learning.
In the federal model of marketing scenario-to-store forecasts, multiple store attempts involving clues to the store, i.e., for a potential customer, there may be multiple cities and multiple stores may meet the customer's needs. Behavior prediction modes in the related art need to manage a large number of models, so that inconvenience is brought to store management and server prediction; moreover, a large number of models need to be called during prediction, so that serious burden is brought to a system; meanwhile, a plurality of prediction results are output by a plurality of models respectively, and further integration and comparison are needed to obtain a target store, so that the prediction efficiency is low.
In order to solve the problems in the related art, embodiments of the present application provide a behavior prediction method, which may consider a user as a purchase-oriented user when the user performs a purchase search on a terminal of the user. By combining with the user portrait and recommending the most appropriate store for the user portrait, the consumption experience of the user can be improved to a great extent, and the sales volume is increased. The following illustrates a scenario of marketing a vehicle.
Before performing behavior prediction, a sample feature set is obtained to train a behavior prediction model. The embodiment of the application is based on the sample feature set formed by obtaining the sample features of the first participant and the second participant through federal learning (such as longitudinal federal learning), and can ensure the security of privacy information of each participant.
Step S302, receiving a behavior prediction request sent by a user terminal.
When a user searches for and browses a certain brand of vehicle on the browser of his terminal, the user is considered to be a user who may purchase a vehicle. The terminal sends a behavior prediction request to the behavior prediction device so as to show stores meeting the vehicle purchasing requirements of the user.
Step S303, obtaining the user characteristics of the user to be predicted based on the behavior prediction request.
After receiving a behavior prediction request sent by a user terminal, the behavior prediction equipment analyzes the prediction request to obtain identification information of a user to be predicted, and determines user characteristics of the user to be predicted based on the identification information.
In an embodiment of the present application, the user characteristics include at least one of: search stores, search vehicle models, age, gender, city of life, major activity areas, purchasing power determined based on historical consumption habits, shopping preferences determined based on historical shopping information, and the like.
Step S304, inputting the user characteristics into the trained behavior prediction model to obtain at least one prediction result.
Here, the prediction result includes a probability that the user to be predicted goes to a destination to be predicted.
In the embodiment of the application, before behavior prediction is performed, the behavior prediction device trains the constructed behavior prediction model in advance, namely the behavior prediction model which is trained well based on federal learning. After the user characteristics of the user to be predicted are determined, the user characteristics are input into the trained behavior prediction model, and the shop-to-shop willingness of the user to go to each shop can be obtained. Here, the arrival willingness may be an arrival probability. The store that matches the user characteristics, i.e., the in-store vehicle meets the user's needs, the greater the user's willingness to go to the store, i.e., the greater the user's probability of arriving at the store.
And after the user characteristics are determined, inputting the user characteristics into the trained behavior prediction model to obtain at least one prediction result, namely the shop-arriving intention of the user to go to each store.
Step S305, sending the at least one prediction result to the user terminal, so that the user terminal outputs the at least one prediction result.
And after the desire to go to each store is calculated based on the behavior prediction model, returning the calculation result to the user terminal so that the user can determine the prediction result of each store according to the prediction result output by the terminal.
The user terminal according to the embodiment of the present application may be a terminal of a user who needs to purchase a vehicle, and the user of the purchaser may determine which store the user goes to buy the vehicle that best meets his or her own requirements according to the prediction result. The user terminal indicated by the embodiment of the application can also be a terminal of a seller selling the vehicles, and the seller user can determine which store the buyer user goes to most possibly buy the vehicle meeting the requirements of the seller user according to the prediction result, so that a more appropriate store can be pertinently recommended to the user, the purchasing experience of the user is improved, and the sales volume is increased.
The behavior prediction method provided by the embodiment of the application is applied to a first party of federal learning, and comprises the following steps: acquiring a sample feature set, inputting the sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model, wherein the behavior prediction model is a model based on federal learning training; receiving a behavior prediction request sent by a user terminal; acquiring user characteristics of a user to be predicted based on the behavior prediction request; inputting the user characteristics into a trained behavior prediction model to obtain at least one prediction result, wherein the prediction result comprises the probability that the user to be predicted goes to a destination to be predicted; and sending the at least one prediction result to the user terminal so that the user terminal outputs the at least one prediction result to realize the prediction of the behavior intention of the user to be predicted, and when the prediction result is applied to a marketing scene, a more appropriate store can be recommended to the user in a targeted manner, the purchasing experience of the user is improved, and the sales volume is increased.
In some embodiments, step S301 "of the embodiment shown in fig. 3, acquiring a sample feature set, and inputting the sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model", may be implemented by the following steps:
step S3011, obtain at least one first feature of the first party.
Here, the first feature is a feature of the first party itself.
In step S3012, a public key for homomorphic encryption and a private key for homomorphic decryption are generated.
In the embodiment of the application, the behavior prediction device generates the public key and the private key for homomorphic encryption and decryption, so that each participant does not need to send original data of the participant to other participants, and the privacy of the original data of each participant can be protected.
Homomorphic encryption is a cryptographic technique based on the theory of computational complexity of mathematical problems. The homomorphic encrypted data is processed to produce an output, which is decrypted, the result being the same as the output obtained by processing the unencrypted original data in the same way. The generated first public key and the first private key can be used for any homomorphic encryption of addition homomorphic encryption, multiplication homomorphic encryption, mixed multiplication homomorphic encryption, subtraction homomorphic encryption, division homomorphic encryption, algebraic homomorphic encryption (also called fully homomorphic encryption) and arithmetic homomorphic encryption. Here, the fully homomorphic encryption means that the encryption function satisfies both the addition homomorphism and the multiplication homomorphy.
In some embodiments, the behavior prediction device may generate a public key and a private key for additive homomorphic encryption and decryption, or the behavior prediction device may generate a public key and a private key for multiplicative homomorphic encryption and decryption, or the behavior prediction device may generate a public key and a private key for fully homomorphic encryption and decryption. Compared with the generation of the public key and the private key for the full homomorphic encryption and decryption, the generation of the public key and the private key for the addition homomorphic encryption and decryption can improve the operation efficiency.
Step S3013, perform feature alignment on the at least one first feature and the at least one second feature of the second party based on the public key and the private key, to obtain a sample feature set.
Training based on a SecureBoost model can be divided into two steps, wherein the first step is to perform data alignment to obtain a sample characteristic set; the second step is to construct a Boost tree, i.e., the behavior prediction model constructed in step S13.
The difficulty of data alignment is how to make private information realize data alignment without being exposed, for example, participant a has XA: { X1, X2, X3, X4} four data, party B has XB: { x1, x2, x3, x5} if party a and party B are allowed to find the union { x1, x2, x3} without knowing the data of each other.
In the embodiment of the present application, the implementation manner of determining the sample characteristics may be: the behavior prediction device may first determine at least one second encoded value based on the public key, the private key, and at least one second characteristic of the second participant, while determining at least one first encoded value based on the at least one first characteristic. The at least one first encoded value and the at least one second encoded value are then sent to the second participant to cause the second participant to determine a feature intersection based on the at least one first encoded value and the at least one second encoded value. Finally, the behavior prediction device determines a sample feature set based on the feature collection and the at least one first encoding value.
And S3014, inputting the sample feature set into a behavior prediction model constructed based on federal learning for training to obtain a trained behavior prediction model.
According to the embodiment of the application, the private key and the public key are generated through homomorphic encryption and decryption, and the constructed behavior prediction model can be trained on the premise that the original data of the first participant and the original data of the second participant are not leaked, so that the trained behavior prediction model is obtained.
In some embodiments, the above step S3013 "performing feature alignment on the at least one first feature and the at least one second feature of the second participant based on the public key and the private key to obtain the sample feature set" may be implemented by:
step S131, determining each second code value corresponding to each second feature based on the public key, the private key, and each second feature.
Determining the respective second encoding values may be implemented as: the behavior prediction device sends the public key to a second participant. And the second participant encrypts each second feature based on the public key to obtain each corresponding second ciphertext feature, then the second participant sends each second ciphertext feature to the behavior prediction equipment, and the behavior prediction equipment decrypts each second ciphertext feature according to the private key to obtain each second encoding value corresponding to each second feature.
Step S132, determining, based on each first feature, each first code value corresponding to the each first feature.
The behavior prediction device determining the at least one first coded value may be implemented as: respectively encoding each first feature to obtain each first feature encoding value corresponding to each first feature, and then decrypting each first feature encoding value according to the private key to obtain each first feature decryption value; and coding each first feature decryption value to obtain each first coding value corresponding to each first feature.
In this embodiment of the present application, the encoding manner may be hash encoding.
Step S133, sending each first encoded value and each second encoded value to the second participant.
The behavior prediction device sends the determined at least one first coded value and at least one second coded value to a second participant, the second participant codes the at least one second coded value to obtain at least one second feature coded value, and the intersection of the at least one second feature coded value and the at least one first coded value is obtained to obtain feature intersection. And the second party sends the obtained feature intersection to the behavior prediction equipment.
Step S134, receiving a feature intersection sent by the second participant, where the feature intersection is determined based on the first encoding values and the second encoding values.
Step S135, determining the feature intersection and the intersection of the first encoding values as the sample feature set.
And the behavior prediction equipment determines the intersection of the feature intersection and each first coding value as a sample feature set by comparing each first coding value with the feature intersection.
In some embodiments, the behavior prediction device may also send the sample feature set to the second participant.
According to the embodiment of the application, data alignment can be achieved on the premise that original data of all participants are not leaked and privacy information of all participants is not exposed, and sample characteristics are obtained.
In the embodiment of the application, the granularity of the acquired user features is related to the number of the acquired prediction results. Because one store predicts to obtain a prediction result, generally, for the same user to be predicted, the larger the granularity is, the more the number of the obtained prediction results is, that is, the larger the prediction range is, the more shopping stores can be selected by the user; conversely, the smaller the granularity, the smaller the number of prediction results obtained, i.e., the smaller the prediction range, the fewer shopping malls the user can select. When the granularity is not limited, the user can shop at any one of all stores; when the granularity is limited to "city," the user can shop at any store within the city; when the granularity is limited to 'stores', the prediction result is only one, and the user can only shop at the stores.
Based on this, when the user features are divided into three granularities, according to the granularity of the user features, the step S303 can be implemented at least in the following three ways:
the first mode is as follows: only the user is limited, and when the granularity is not limited, the step S303 "obtaining the user characteristics of the user to be predicted based on the behavior prediction request" may be implemented as:
step S303a1, analyzing the prediction request to obtain the identification information of the user to be predicted.
Step S303a2, determining the identification information as the user characteristic of the user to be predicted.
In a second manner, when defining the region granularity, the step S303 "obtaining the user characteristics of the user to be predicted based on the behavior prediction request" may be implemented as:
step S303b1, analyzing the prediction request to obtain the identification information of the user to be predicted.
Step S303b2, obtaining at least one region to be predicted according to the identification information.
Step S303b3, determining the identification information and the at least one region to be predicted as at least one user feature of the user to be predicted.
In a third manner, when defining the destination granularity, the step S303 "obtaining the user characteristics of the user to be predicted based on the behavior prediction request" may be implemented as:
step S303c1, analyzing the prediction request to obtain the identification information of the user to be predicted.
Step S303c2, obtaining at least one region to be predicted according to the identification information.
Step S303c3, obtaining at least one destination to be predicted according to the at least one area to be predicted.
Step S303c4, determining the identification information, the at least one area to be predicted and the at least one destination to be predicted as at least one user characteristic of the user to be predicted.
Table 1 shows an example of the prediction result of the user U1 to be predicted, and as shown in table 1, the user characteristic obtained in the first manner is U1, and the prediction results obtained at this time are P1 to P5, the user characteristic obtained in the second manner is (U1, city a), and the prediction results obtained at this time are P1 to P3, and the user characteristic obtained in the third manner is (U1, city a, store 2) and the prediction result obtained is P2.
Table 1 example of prediction result of user U1 to be predicted
In the embodiment of the application, the prediction results of diversity can be output by changing the granularity of the user characteristics.
Based on the foregoing embodiments, an embodiment of the present application further provides a behavior prediction method, and fig. 4 is a schematic diagram of another implementation flow of the behavior prediction method provided in the embodiment of the present application, as shown in fig. 4, the behavior prediction method includes the following steps:
step S401, a sample feature set is obtained and input into a pre-constructed behavior prediction model for training, and a trained behavior prediction model is obtained.
Here, the behavior prediction model is a model trained based on federal learning.
Step S402, receiving a behavior prediction request sent by a user terminal.
Step S403, obtaining the user characteristics of the user to be predicted based on the behavior prediction request.
Step S404, inputting the user characteristics into the trained behavior prediction model to obtain at least one prediction result.
Here, the prediction result includes a probability that the user to be predicted goes to a destination to be predicted.
Steps S401 to S404 correspond to steps S301 to S304, respectively, and the implementation manner thereof is described in steps S301 to S304.
And S405, sequencing the at least one prediction result according to a preset sequencing mode to obtain a sequenced prediction result.
Step S406, sending the ordered prediction result to the user terminal, so that the user terminal outputs the ordered prediction result.
When the prediction results are multiple, the multiple prediction results can be sorted according to the probability of the user to be predicted going to multiple destinations to be predicted, for example, sorted from large to small or sorted from small to large, so that the ordered prediction results are obtained, and the user terminal outputs the ordered prediction results, so that the user can conveniently view the ordered prediction results.
Based on the foregoing embodiments, an embodiment of the present application further provides a behavior prediction method, which is applied to the network architecture shown in fig. 1, fig. 5 is a schematic diagram of a further implementation flow of the behavior prediction method provided in the embodiment of the present application, and as shown in fig. 5, the behavior prediction method includes the following steps:
in step S501, the behavior prediction device obtains at least one first feature of the first participant.
In step S502, the behavior prediction apparatus generates a public key for homomorphic encryption and a private key for homomorphic decryption.
Step S503, the behavior prediction device sends the public key to the server.
Step S504, the server side encrypts each second characteristic based on the public key to obtain each second ciphertext characteristic.
And step S505, the server side sends the second ciphertext characteristics to behavior prediction equipment.
Step S506, the behavior prediction device decrypts each second ciphertext feature according to the private key to obtain each second encoded value corresponding to each second feature.
Step S507, the behavior prediction device encodes each first feature to obtain each first feature encoded value corresponding to each first feature.
Step S508, the behavior prediction device decrypts the first feature code values according to the private key to obtain first feature decrypted values.
Step S509, the behavior prediction device encodes each first feature decryption value to obtain each first encoded value corresponding to each first feature.
Step S510, the behavior prediction device sends each first encoded value and each second encoded value to the server.
And step S511, the server side encodes each second coded value to obtain each second characteristic coded value.
Step S512, the server obtains a feature intersection of each first encoded value and each second encoded value.
Step S513, the server sends the feature intersection to the behavior prediction device.
In step S514, the behavior prediction device determines the feature intersection and the intersection of the first encoding values as the sample feature set.
And step S515, inputting the sample feature set into a behavior prediction model constructed based on federal learning by the behavior prediction equipment for training to obtain a trained behavior prediction model.
Here, the behavior prediction model is a model trained based on federal learning.
In step S516, the behavior prediction device receives a behavior prediction request sent by the user terminal.
And step S517, the behavior prediction equipment acquires the user characteristics of the user to be predicted based on the behavior prediction request.
Step S518, the behavior prediction device inputs the user characteristics to the trained behavior prediction model to obtain at least one prediction result.
Here, the prediction result includes a probability that the user to be predicted goes to a destination to be predicted.
Step S519, the behavior prediction device sends the at least one prediction result to the user terminal.
Step S520, the ue outputs the at least one prediction result.
In the behavior prediction method provided by the embodiment of the application, a behavior prediction device firstly obtains a sample feature set, inputs the sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model, and the behavior prediction model is a model based on federal learning training; after receiving a behavior prediction request sent by a user terminal; acquiring user characteristics of a user to be predicted based on the behavior prediction request; inputting the user characteristics into a trained behavior prediction model to obtain at least one prediction result, wherein the prediction result comprises the probability that the user to be predicted goes to a destination to be predicted; and sending the at least one prediction result to the user terminal so that the user terminal outputs the at least one prediction result to realize the prediction of the behavior intention of the user to be predicted, and when the prediction result is applied to a marketing scene, a more appropriate store can be recommended to the user in a targeted manner, the purchasing experience of the user is improved, and the sales volume is increased.
Next, an exemplary application of the embodiment of the present application in a practical application scenario will be described.
The nature of the model requirements is a translation prediction to store intent for customer clues (car purchase intent). Typically, dealers obtain clue customers who may have potential car buying intentions through some channels, and by tracking and visiting the customers, records of whether the customers finally arrive at the store for viewing can be obtained. Based on the clues, with the store-to-store records as labels and more data provided by the data source as features, a model is trained to predict whether other clue customers arrive at the store.
In the model of marketing scenario-to-store forecasting, there are multiple store attempts involving clue customer-to-store, i.e., for a potential customer, there may be multiple cities, and multiple stores may meet the customer's needs. According to the scheme in the related technology, a model needs to be trained for each store, the server side needs to predict the prediction result (score) of whether a customer arrives at the store according to the model of each store, compare a plurality of prediction scores, and select the store with the highest score to perform clue store distribution. The method needs to manage a large number of models, and brings inconvenience to store management and server prediction; moreover, a large number of models need to be called during prediction, so that serious burden is brought to a system; meanwhile, the final prediction results are output by the multiple models respectively, and further integration and comparison are needed.
In the prediction process of the related art, the prediction is performed only once for one client identification number (ID). The client (Guest) (i.e. the user terminal in the above) sends a prediction request, and the two parties (including the first party and the second party in the above) respectively calculate scores and then aggregate the scores and return the scores at the Guest side. In the scheme, for the same ID, the prediction is carried out for a plurality of times according to the city and store needing to be predicted, the Host side (Host) (namely the behavior prediction device in the above) is only calculated once, and the Guest side can carry out the prediction for a plurality of times in the model according to different conditions and simultaneously returns the corresponding score values in sequence.
In the embodiment of the present application, due to the change of the online module, the training module also needs to perform corresponding adjustment and modification: in the training process, city and store information needs to be introduced, and simultaneously, the city and store information is input into a training model for training.
In the conventional prediction process, prediction occurs only once for one ID. And (4) sending a prediction request through the Guest party, and summarizing the prediction request on the Guest side after the two parties calculate scores respectively.
In the embodiment of the application, for the same ID, the city and the store which are predicted according to needs are predicted for several times, the Host side is only calculated once, the guest side can predict for several times in the model according to different conditions, and corresponding score values are returned sequentially.
Fig. 6 is a schematic diagram of a model training phase provided in the embodiment of the present application, referring to fig. 6, in the training phase, a feature "city" and a feature "store" in a client 601 are used as virtual variables, a feature included in the client 601 and a feature included in a host 602 are obtained by a behavior prediction model 603 based on a Secureboost model training of federal learning.
Fig. 7 is a schematic diagram of the prediction phase provided in the embodiment of the present application, and referring to fig. 7, a customer with an ID of U1 predicts the arrival probability of the customer at store 1 to store 5 in city a and city B. According to the traditional mode, 5 different models need to be trained and managed, and when the models are called online, the 5 models need to be called respectively for operation, and results are integrated manually. In the embodiment of the present application, the corresponding arrival willingness P1 to P5 can be obtained only by bringing the corresponding characteristics, i.e., (city, store) in the prediction module. In the prediction, the client 701 has a list of the combinations of cities and stores that the client intends to make, obtains the arrival willingness of a plurality of combinations of cities and stores (corresponding to the above prediction results), and allows the user to select the combination of cities and stores having the highest probability as the target store.
The prediction method provided by the embodiment of the application can reduce the number of online models and reduce the maintenance time and cost of model management. When the online calling is carried out, the corresponding model of the store does not need to be matched manually, the model can directly return the probability value calculated according to the condition, and therefore the suitable store is selected.
Continuing with the exemplary structure of the behavior prediction apparatus provided by the embodiment of the present application as a software module, in some embodiments, as shown in fig. 2, the software module stored in the behavior prediction apparatus 80 of the memory 140 may include:
the training module 81 is used for acquiring a sample feature set, inputting the sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model, wherein the behavior prediction model is a model based on federal learning training;
a receiving module 82, configured to receive a behavior prediction request sent by a user terminal;
a first obtaining module 83, configured to obtain a user characteristic of the user to be predicted based on the behavior prediction request;
an input module 84, configured to input the user characteristics into a trained behavior prediction model to obtain at least one prediction result, where the prediction result includes a probability that the user to be predicted goes to a destination to be predicted;
a sending module 85, configured to send the at least one prediction result to the user terminal, so that the user terminal outputs the at least one prediction result.
In some embodiments, the training module 81 is further configured to:
obtaining at least one first characteristic of the first party;
generating a public key for homomorphic encryption and a private key for homomorphic decryption;
performing feature alignment on the at least one first feature and at least one second feature of a second party based on the public key and the private key to obtain a sample feature set;
and inputting the sample feature set into a behavior prediction model constructed based on federal learning for training to obtain a trained behavior prediction model.
In some embodiments, the training module 81 is further configured to:
determining each second code value corresponding to each second feature based on the public key, the private key and each second feature;
determining, based on each first feature, each first encoded value corresponding to the each first feature;
sending the respective first encoded values and the respective second encoded values to the second participant;
receiving a feature intersection sent by the second participant, the feature intersection being determined based on the respective first encoding values and the respective second encoding values;
and determining the feature intersection and the intersection of the feature intersection and the various first coding values as the sample feature set.
In some embodiments, the training module 81 is further configured to:
sending the public key to a second party;
receiving each second ciphertext feature sent by the second party, wherein the second ciphertext feature is obtained by encrypting the second feature by the second party based on the public key;
and decrypting each second ciphertext characteristic according to the private key to obtain each second coding value corresponding to each second characteristic.
In some embodiments, the training module 81 is further configured to:
respectively coding each first characteristic to obtain each first characteristic coding value corresponding to each first characteristic;
decrypting each first characteristic coding value according to the private key to obtain each first characteristic decrypted value;
and coding each first feature decryption value to obtain each first coding value corresponding to each first feature.
In some embodiments, the first obtaining module 83 is further configured to:
analyzing the prediction request to obtain the identification information of the user to be predicted;
determining the identification information as the user characteristics of the user to be predicted;
alternatively, the first and second electrodes may be,
analyzing the prediction request to obtain the identification information of the user to be predicted;
acquiring at least one region to be predicted according to the identification information;
determining the identification information and the at least one region to be predicted as at least one user characteristic of the user to be predicted;
alternatively, the first and second electrodes may be,
analyzing the prediction request to obtain the identification information of the user to be predicted;
acquiring at least one region to be predicted according to the identification information;
acquiring at least one destination to be predicted according to the at least one area to be predicted;
determining the identification information, the at least one area to be predicted and the at least one destination to be predicted as at least one user characteristic of the user to be predicted.
In some embodiments, the behavior prediction device 80 further includes:
the sequencing module is used for sequencing the at least one prediction result according to a preset sequencing mode to obtain an ordered prediction result;
correspondingly, the sending module 85 is further configured to:
and sending the ordered prediction result to the user terminal so that the user terminal outputs the ordered prediction result.
Here, it should be noted that: the above description of the embodiment of the behavior prediction apparatus is similar to the above description of the method, and has the same advantageous effects as the embodiment of the method. For technical details not disclosed in the embodiments of the behavior prediction device of the present application, those skilled in the art should understand with reference to the description of the embodiments of the method of the present application.
Embodiments of the present application provide a storage medium having stored therein executable instructions, which when executed by a processor, will cause the processor to perform the methods provided by embodiments of the present application, for example, the methods as illustrated in fig. 3 to 5.
In some embodiments, the storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EE PROM, flash, magnetic surface memory, optical disk, or CD-ROM; or may be various devices including one or any combination of the above memories.
In some embodiments, executable instructions may be written in any form of programming language (including compiled or interpreted languages), in the form of programs, software modules, scripts or code, and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.
By way of example, executable instructions may correspond, but do not necessarily have to correspond, to files in a file system, and may be stored in a portion of a file that holds other programs or data, such as in one or more scripts in a hypertext Markup Language (H TML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub-programs, or portions of code).
By way of example, executable instructions may be deployed to be executed on one computing device or on multiple computing devices at one site or distributed across multiple sites and interconnected by a communication network.
The above description is only an example of the present application, and is not intended to limit the scope of the present application. Any modification, equivalent replacement, and improvement made within the spirit and scope of the present application are included in the protection scope of the present application.
Claims (11)
1. A method of behavioral prediction for a first party in federal learning, the method comprising:
acquiring a sample feature set, inputting the sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model, wherein the behavior prediction model is a model based on federal learning training;
receiving a behavior prediction request sent by a user terminal;
acquiring user characteristics of a user to be predicted based on the behavior prediction request;
inputting the user characteristics into a trained behavior prediction model to obtain at least one prediction result, wherein the prediction result comprises the probability that the user to be predicted goes to a destination to be predicted;
and sending the at least one prediction result to the user terminal so that the user terminal outputs the at least one prediction result.
2. The method according to claim 1, wherein the obtaining a sample feature set and inputting the sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model comprises:
obtaining at least one first characteristic of the first party;
generating a public key for homomorphic encryption and a private key for homomorphic decryption;
performing feature alignment on the at least one first feature and at least one second feature of a second party based on the public key and the private key to obtain a sample feature set;
and inputting the sample feature set into a behavior prediction model constructed based on federal learning for training to obtain a trained behavior prediction model.
3. The method of claim 2, wherein the feature aligning the at least one first feature and the at least one second feature of the second participant based on the public key and the private key to obtain a sample feature set comprises:
determining each second code value corresponding to each second feature based on the public key, the private key and each second feature;
determining, based on each first feature, each first encoded value corresponding to the each first feature;
sending the respective first encoded values and the respective second encoded values to the second participant;
receiving a feature intersection sent by the second participant, the feature intersection being determined based on the respective first encoding values and the respective second encoding values;
and determining the feature intersection and the intersection of the feature intersection and the various first coding values as the sample feature set.
4. The method of claim 3, wherein determining, based on the public key, the private key, and the respective second features, respective second encoded values corresponding to the respective second features comprises:
sending the public key to a second party;
receiving each second ciphertext feature sent by the second party, wherein the second ciphertext feature is obtained by encrypting the second feature by the second party based on the public key;
and decrypting each second ciphertext characteristic according to the private key to obtain each second coding value corresponding to each second characteristic.
5. The method according to claim 3, wherein determining, based on the respective first features, respective first encoded values corresponding to the respective first features comprises:
respectively coding each first characteristic to obtain each first characteristic coding value corresponding to each first characteristic;
decrypting each first characteristic coding value according to the private key to obtain each first characteristic decrypted value;
and coding each first feature decryption value to obtain each first coding value corresponding to each first feature.
6. The method of claim 1, wherein the obtaining the user characteristics of the user to be predicted based on the behavior prediction request comprises:
analyzing the prediction request to obtain the identification information of the user to be predicted;
determining the identification information as the user characteristics of the user to be predicted;
alternatively, the first and second electrodes may be,
analyzing the prediction request to obtain the identification information of the user to be predicted;
acquiring at least one region to be predicted according to the identification information;
determining the identification information and the at least one region to be predicted as at least one user characteristic of the user to be predicted;
alternatively, the first and second electrodes may be,
analyzing the prediction request to obtain the identification information of the user to be predicted;
acquiring at least one region to be predicted according to the identification information;
acquiring at least one destination to be predicted according to the at least one area to be predicted;
determining the identification information, the at least one area to be predicted and the at least one destination to be predicted as at least one user characteristic of the user to be predicted.
7. The method according to any one of claims 1 to 6, further comprising:
sequencing the at least one prediction result according to a preset sequencing mode to obtain a sequential prediction result;
correspondingly, the sending the at least one prediction result to the user terminal to enable the user terminal to output the at least one prediction result includes:
and sending the ordered prediction result to the user terminal so that the user terminal outputs the ordered prediction result.
8. A behavior prediction apparatus for use with a first party to federal learning, the apparatus comprising:
the training module is used for acquiring a sample feature set, inputting the sample feature set into a pre-constructed behavior prediction model for training to obtain a trained behavior prediction model, and the behavior prediction model is a model based on federal learning training;
the receiving module is used for receiving a behavior prediction request sent by a user terminal;
the first obtaining module is used for obtaining the user characteristics of the user to be predicted based on the behavior prediction request;
the input module is used for inputting the user characteristics into a trained behavior prediction model to obtain at least one prediction result, and the prediction result comprises the probability that the user to be predicted goes to a destination to be predicted;
and the sending module is used for sending the at least one prediction result to the user terminal so as to enable the user terminal to output the at least one prediction result.
9. A behavior prediction device for application to a first party of federal learning, the device comprising:
a memory for storing executable instructions;
a processor for implementing the method of any one of claims 1 to 7 when executing executable instructions stored in the memory.
10. A computer-readable storage medium having stored thereon executable instructions for causing a processor, when executed, to implement the method of any one of claims 1 to 7.
11. A computer program product comprising a computer program, characterized in that the computer program realizes the method of any of claims 1 to 7 when executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011644642.7A CN112700281A (en) | 2020-12-30 | 2020-12-30 | Behavior prediction method, behavior prediction device, behavior prediction equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011644642.7A CN112700281A (en) | 2020-12-30 | 2020-12-30 | Behavior prediction method, behavior prediction device, behavior prediction equipment and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112700281A true CN112700281A (en) | 2021-04-23 |
Family
ID=75514279
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011644642.7A Pending CN112700281A (en) | 2020-12-30 | 2020-12-30 | Behavior prediction method, behavior prediction device, behavior prediction equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112700281A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111340565A (en) * | 2020-03-20 | 2020-06-26 | 北京爱笔科技有限公司 | Information recommendation method, device, equipment and storage medium |
| CN113537633A (en) * | 2021-08-09 | 2021-10-22 | 中国电信股份有限公司 | Prediction method, device, equipment, medium and system based on longitudinal federal learning |
| CN113822709A (en) * | 2021-09-15 | 2021-12-21 | 摩拜(北京)信息技术有限公司 | Travel data processing method and device and server |
-
2020
- 2020-12-30 CN CN202011644642.7A patent/CN112700281A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111340565A (en) * | 2020-03-20 | 2020-06-26 | 北京爱笔科技有限公司 | Information recommendation method, device, equipment and storage medium |
| CN113537633A (en) * | 2021-08-09 | 2021-10-22 | 中国电信股份有限公司 | Prediction method, device, equipment, medium and system based on longitudinal federal learning |
| CN113822709A (en) * | 2021-09-15 | 2021-12-21 | 摩拜(北京)信息技术有限公司 | Travel data processing method and device and server |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112700281A (en) | Behavior prediction method, behavior prediction device, behavior prediction equipment and computer readable storage medium | |
| CN110189192B (en) | Information recommendation model generation method and device | |
| Xiao et al. | SRA: Secure reverse auction for task assignment in spatial crowdsourcing | |
| CN109033865B (en) | Task allocation method for privacy protection in space crowdsourcing | |
| JP7095140B2 (en) | Multi-model training methods and equipment based on feature extraction, electronic devices and media | |
| CN112132198B (en) | Data processing method, device and system and server | |
| CN113537633B (en) | Prediction method, device, equipment, medium and system based on longitudinal federal learning | |
| CN112182644A (en) | Data processing method and device and electronic equipment | |
| CN111666460A (en) | User portrait generation method and device based on privacy protection and storage medium | |
| CN110851869A (en) | Sensitive information processing method and device and readable storage medium | |
| CN112347500B (en) | Machine learning method, device, system, equipment and storage medium of distributed system | |
| CN113032840A (en) | Data processing method, device, equipment and computer readable storage medium | |
| CN110969264B (en) | Model training method, distributed prediction method and system thereof | |
| JP2007501975A (en) | Data processing system and method | |
| CN113542228B (en) | Data transmission method and device based on federal learning and readable storage medium | |
| CN109658130A (en) | It is recycled based on block chain and provides Coupon Method, device, equipment and storage medium | |
| CN111027981B (en) | Method and device for multi-party joint training of risk assessment model for IoT (Internet of things) machine | |
| Ying et al. | PrivacySignal: Privacy-preserving traffic signal control for intelligent transportation system | |
| CN114611008B (en) | User service strategy determination method and device based on federal learning and electronic equipment | |
| CN111709051A (en) | Data processing method, device and system, computer storage medium and electronic equipment | |
| CN110209994A (en) | Matrix decomposition recommendation method based on homomorphic cryptography | |
| CN111490995A (en) | Model training method and device for protecting privacy, data processing method and server | |
| CN110610098A (en) | Data set generation method and device | |
| CN114978484A (en) | Data processing method and device for protecting privacy and computer equipment | |
| CN109690516A (en) | A kind of software-based switch for providing a user product and/or servicing without damaging its privacy |
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 |