CN117421485A - Message pushing method, device and storage medium - Google Patents
Message pushing method, device and storage medium Download PDFInfo
- Publication number
- CN117421485A CN117421485A CN202311609910.5A CN202311609910A CN117421485A CN 117421485 A CN117421485 A CN 117421485A CN 202311609910 A CN202311609910 A CN 202311609910A CN 117421485 A CN117421485 A CN 117421485A
- Authority
- CN
- China
- Prior art keywords
- target user
- displayed
- recommendation system
- sampling period
- functional area
- 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 64
- 238000005070 sampling Methods 0.000 claims abstract description 94
- 230000006399 behavior Effects 0.000 claims description 94
- 238000012360 testing method Methods 0.000 claims description 26
- 238000012549 training Methods 0.000 claims description 23
- 230000000737 periodic effect Effects 0.000 claims description 20
- 230000015654 memory Effects 0.000 claims description 14
- 230000003442 weekly effect Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 7
- 230000006870 function Effects 0.000 claims description 7
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- 238000013112 stability test Methods 0.000 claims description 2
- 238000012353 t test Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract description 3
- 238000012545 processing Methods 0.000 description 10
- 238000004891 communication Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 230000000694 effects Effects 0.000 description 5
- 230000007246 mechanism Effects 0.000 description 5
- 230000001932 seasonal effect Effects 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000003542 behavioural effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000005295 random walk Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 238000000780 augmented Dickey–Fuller test Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- YHXISWVBGDMDLQ-UHFFFAOYSA-N moclobemide Chemical compound C1=CC(Cl)=CC=C1C(=O)NCCN1CCOCC1 YHXISWVBGDMDLQ-UHFFFAOYSA-N 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/02—Banking, e.g. interest calculation or account maintenance
-
- 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Technology Law (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present disclosure relates to the field of big data technology, and may be used in the field of finance, in particular, to a method, an apparatus and a storage medium for pushing a message, including: acquiring user information of a target user; extracting a prediction model which corresponds to a target user and is trained, and historical behavior data in a sampling period; the historical behavior data comprises operation module data, operation button data, operation starting time and operation ending time; obtaining average stay time of a target user for each functional area of a recommendation system in a sampling period according to the sampling period and the historical behavior data, wherein the functional areas comprise a module area and an operation area; the average residence time of each functional area is imported into a prediction model to obtain the content to be displayed of the target user in the recommendation system, wherein the displayed content comprises each functional area in the recommendation system, parameter change of the recommendation system caused by user false touch can be eliminated, and the recommendation system can recommend more accurate displayed content for the target system.
Description
Technical Field
The invention relates to the technical field of big data, and can be used in the financial field, in particular to a message pushing method, a message pushing device and a storage medium.
Background
Currently, message pushing has been widely used in new media. The method relies on the modern information technology to accurately push personalized information to audience groups, so that the cost is effectively saved, the time for screening information by users is reduced, and the user experience is improved. However, most of the message pushing technologies only stay at the step of collecting user behaviors, and after pages or modules operated by the user are recorded, data related to the records are put into the user. This approach, while quite simple and straightforward, has drawbacks: it is impossible to distinguish whether the user's historical behavior is truly behavior or careless mishandling behavior. If the misoperation behavior data is recorded and the information is pushed based on the misoperation behavior data of the user, the user experience is affected to a certain extent and unnecessary burden of the system is increased.
Disclosure of Invention
In view of the foregoing problems in the prior art, an object of the present disclosure is to provide a method, an apparatus, and a storage medium for pushing a message, so as to solve the problem that in the prior art, information is pushed based on user misoperation behavior data, which will have a certain influence on user experience and increase unnecessary burden of a system.
In order to solve the technical problems, the specific technical scheme is as follows:
in one aspect, provided herein is a message pushing method, including:
responding to the login of a target user to a recommendation system, and acquiring user information of the target user;
extracting a prediction model which is trained and completed and corresponds to the target user and historical behavior data in a sampling period according to the user information; the historical behavior data comprise operation module data, operation button data, operation starting time and operation ending time;
obtaining average residence time of the target user in the sampling period for each functional area of the recommendation system according to the historical behavior data, wherein the functional areas comprise a module area and an operation area;
and importing the corresponding numbers of the functional areas and the average residence time of the functional areas into the prediction model to obtain the content to be displayed of the target user in the recommendation system, wherein the display content comprises the functional areas in the recommendation system, and the prediction model is trained by using the historical behavior data.
As one embodiment herein, the sampling period comprises hourly, daily, or weekly; the prediction model comprises a first prediction model, a second prediction model and a third prediction model;
The step of importing the average residence time of each functional area into the prediction model to obtain the content to be displayed of the target user in the recommendation system, further comprises the steps of:
the average residence time of each functional area calculated according to each hour is imported into a first prediction model to obtain a first sub-set to be displayed of the target user in the recommendation system;
the average residence time of each functional area obtained through daily calculation is imported into a second prediction model, and a second sub-set to be displayed of the target user in the recommendation system is obtained;
the average residence time of each functional area obtained through weekly calculation is imported into a third prediction model, and a third sub-set to be displayed of the target user in the recommendation system is obtained;
and fusing the first sub-set to be displayed, the second sub-set to be displayed and the third sub-set to be displayed to obtain the content to be displayed of the target user in the recommendation system.
As an embodiment herein, the fusing the first sub-to-be-displayed set, the second sub-to-be-displayed set, and the third sub-to-be-displayed set to obtain to-be-displayed content of the target user in the recommendation system further includes:
Determining whether the same display content exists in the first sub-set to be displayed, the second sub-set to be displayed and the third sub-set to be displayed, and if so, displaying the display content in a primary area in the recommendation system;
determining whether the same display content exists in any two sub-to-be-displayed sets of the first sub-to-be-displayed set, the second sub-to-be-displayed set and the third sub-to-be-displayed set, and if so, displaying the display content in a secondary area in the recommendation system.
As one embodiment herein, the method for obtaining the user information and the historical behavior data includes:
acquiring the user information and the historical behavior data of the target user to a buried point server through a buried point component which is embedded in advance;
transmitting the user information and the historical behavior data in the embedded point server to a database according to a proxy file; the agent file comprises Kafka information and buried point field information; the Kafka information is used for determining the address, the login account and the login password of the database; the embedded point field information is used for filtering the user information and data fields in the historical behavior data.
As an embodiment of the present disclosure, the obtaining, according to the historical behavior data, an average residence time of the target user in the sampling period for each functional area of the recommendation system further includes:
determining the clicking times of the target user for each functional area in the recommendation system according to the historical behavior data in the sampling period;
determining the time length of each stay of the target user in each functional area according to the time interval of the two functional areas continuously clicked by the target user;
and in the sampling period, determining the average residence time of the functional areas in the sampling period according to the residence time of each functional area and the clicking times of each functional area.
As one embodiment herein, determining the duration of stay of the target user in each functional area according to the time interval of two functional areas continuously clicked by the target user includes: calculating the single stay time t of each functional area according to the following formula ai ：
t ai ＝x 1 -ai 1
Wherein t is ai Representing the single dwell time length x of the target user in the ith dwell function area a in the sampling period 1 Indicating the time ai for recording the buried data when the functional area a jumps to the next functional area 1 The first buried data recording time of the ith stay in the functional area a in a certain hour is represented;
and in the sampling period, determining the average residence time of the functional areas in the sampling period according to the residence time of each functional area and the clicking times of each functional area, and further comprising:
calculating the total stay time T of the target user in the functional area in the sampling period according to the following formula a
Wherein T is a T is the total residence time in the sampling period ai Representing the single stay time length of the target user staying in the functional area for the ith time in the sampling period, wherein i is the ith time, and n is the total stay times in the sampling period;
calculating the average residence time v of the target user in the functional area in the sampling period according to the following formula a
Wherein n is a T is the total number of clicks of the functional area a V is the total dwell time of the functional area in the sampling period a For the average residence time of the target user in the functional area during the sampling period.
As one embodiment herein, the predictive model training method includes:
dividing historical behavior data in different sampling periods into a training set and a testing set in proportion;
respectively importing the training set into a random sequence model and a periodic sequence model, and performing repeated iterative training to obtain a random sequence model and a periodic sequence model, wherein the random sequence model and the periodic sequence model are input by corresponding numbers of all functional areas and average residence time of all the functional areas, and the random sequence model and the periodic sequence model are output by corresponding numbers of all the functional areas at the next moment;
And testing the trained random sequence model and the trained periodic sequence model by using a test set, and taking the sequence model with the smallest difference between the test result and the test set as a prediction model.
In another aspect, there is provided herein a message pushing apparatus, including:
the acquisition unit is used for responding to the login of the target user to the recommendation system and acquiring the user information of the target user;
the extraction unit is used for extracting a prediction model which corresponds to the target user and is trained according to the user information and historical behavior data in a sampling period; the historical behavior data comprise operation module data, operation button data, operation starting time and operation ending time;
the importing unit is used for obtaining average stay time of the target user for each functional area of the recommendation system in the sampling period according to the historical behavior data, wherein the functional areas comprise a module area and an operation area;
the prediction unit is used for importing the corresponding numbers of the functional areas and the average residence time of the functional areas into the prediction model to obtain the content to be displayed of the target user in the recommendation system, wherein the display content comprises the functional areas in the recommendation system, and the prediction model is trained by using the historical behavior data.
In another aspect, there is provided a computer device including a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing any one of the message pushing methods when executing the computer program.
In another aspect, there is provided a computer readable storage medium storing a computer program which, when executed by a processor, implements the message pushing method of any one of the above.
By adopting the technical scheme, the user information of the target user is acquired by responding to the login recommendation system of the target user, so that the user information corresponding to the identity of the target user can be acquired; extracting a prediction model which corresponds to the target user and is trained according to the user information, and historical behavior data in a sampling period; the historical behavior data comprise operation module data, operation button data, operation starting time and operation ending time, and a prediction model corresponding to a target user and past historical behavior data can be obtained; the average stay time of the target user in the sampling period for each functional area of the recommendation system is obtained according to the sampling period and the historical behavior data, wherein the functional area comprises a module area and an operation area, so that the stay time of the target user in a plurality of sampling periods for each operation of different functional areas can be determined, and the average stay time of the target user in the sampling period is further determined through the stay time of each operation; the average residence time of each functional area is imported into the prediction model to obtain the content to be displayed of the target user in the recommendation system, wherein the display content comprises each functional area in the recommendation system, the content to be displayed which is relatively close to the content actually required by the target user can be obtained, and further the user is displayed through the recommendation system.
The foregoing and other objects, features and advantages will be apparent from the following more particular description of preferred embodiments, as illustrated in the accompanying drawings.
Drawings
In order to more clearly illustrate the embodiments herein or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments herein and that other drawings may be obtained according to these drawings without inventive effort to a person skilled in the art.
FIG. 1 illustrates an overall system diagram of a message pushing method of embodiments herein;
FIG. 2 is a schematic diagram illustrating steps of a message pushing method according to an embodiment herein;
FIG. 3 illustrates a schematic diagram of a recommendation system of embodiments herein;
FIG. 4 illustrates a schematic diagram of a predictive model training method of embodiments herein;
FIG. 5 illustrates a message pushing device schematic diagram of an embodiment herein;
FIG. 6 shows a schematic diagram of a computer device of embodiments herein.
Description of the drawings:
101. a terminal;
102. a buried point server;
103. a database;
104. an operation server;
501. An acquisition unit;
502. an extraction unit;
503. an introduction unit;
504. a prediction unit;
602. a computer device;
604. a processor;
606. a memory;
608. a driving mechanism;
610. an input/output module;
612. an input device;
614. an output device;
616. a presentation device;
618. a graphical user interface;
620. a network interface;
622. a communication link;
624. a communication bus.
Detailed Description
The following description of the embodiments of the present disclosure will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the disclosure. All other embodiments, based on the embodiments herein, which a person of ordinary skill in the art would obtain without undue burden, are within the scope of protection herein.
It should be noted that the terms "first," "second," and the like in the description and claims herein and in the foregoing figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, apparatus, article, or device that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or device.
The overall system diagram of a message pushing method shown in fig. 1 includes: a terminal 101, a buried point server 102, a database 103, and an arithmetic server 104;
the terminal 101 is configured to run a recommendation system to display recommended content to a user, upload content (behavior data) that the user browses in the recommendation system, and send the browsed content to a buried server.
The embedded point server 102 is used for regularly collecting content (behavior data) which is usually browsed by a user in a recommendation system.
And the database 103 is used for protecting the behavior data and performing data processing on the behavior data so as to train and predict the prediction model.
The operation server 104 is configured to run a plurality of prediction models, where each user is private with a plurality of prediction models, each prediction model has a number, and a prefix of the number may correspond to a personal account number of the user, for example, the account number of the user a is 1100, and the number prefixes of all the prediction models private with the user a are 1100, so as to distinguish different prediction models, and achieve an effect of preparing pushing for the target user.
It should be noted that, user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, presented data, etc.) referred to in the present application are information and data authorized by the user or sufficiently authorized by each party.
In the prior art, whether the historical behavior of the user is real behavior or careless misoperation behavior cannot be distinguished. If the misoperation behavior data is recorded and the information is pushed based on the misoperation behavior data of the user, the user experience is affected to a certain extent and unnecessary burden of the system is increased.
In order to solve the above-mentioned problems, the embodiments herein provide a message pushing method, which can avoid the problem that a user receives an unexpected pushing result by misoperation, and fig. 2 is a schematic step diagram of a message pushing method provided in the embodiments herein, and the present disclosure provides the method operation steps as described in the embodiment or the flowchart, but may include more or fewer operation steps based on conventional or non-creative labor. The order of steps recited in the embodiments is merely one way of performing the order of steps and does not represent a unique order of execution. When a system or apparatus product in practice is executed, it may be executed sequentially or in parallel according to the method shown in the embodiments or the drawings. As shown in fig. 2, the method may include:
step 201, responding to a target user to log in a recommendation system, and acquiring user information of the target user;
Step 202, extracting a prediction model which corresponds to the target user and is trained according to the user information, and historical behavior data in a sampling period; the historical behavior data comprise operation module data, operation button data, operation starting time and operation ending time;
step 203, obtaining average residence time of the target user in the sampling period for each functional area of the recommendation system according to the sampling period and the historical behavior data, wherein the functional areas comprise a module area and an operation area;
and 204, importing the corresponding numbers of the functional areas and the average residence time of the functional areas into the prediction model to obtain the content to be displayed of the target user in the recommendation system, wherein the display content comprises the functional areas in the recommendation system, and the prediction model is trained by using the historical behavior data.
By adopting the technical scheme, the user information of the target user is acquired by responding to the login recommendation system of the target user, so that the user information corresponding to the identity of the target user can be acquired; extracting a prediction model which corresponds to the target user and is trained according to the user information, and historical behavior data in a sampling period; the historical behavior data comprise operation module data, operation button data, operation starting time and operation ending time, and a prediction model corresponding to a target user and past historical behavior data can be obtained; the average stay time of the target user in the sampling period for each functional area of the recommendation system is obtained according to the sampling period and the historical behavior data, wherein the functional area comprises a module area and an operation area, so that the stay time of the target user in a plurality of sampling periods for each operation of different functional areas can be determined, and the average stay time of the target user in the sampling period is further determined through the stay time of each operation; the average residence time of each functional area is imported into the prediction model to obtain the content to be displayed of the target user in the recommendation system, wherein the display content comprises each functional area in the recommendation system, the content to be displayed which is relatively close to the content actually required by the target user can be obtained, and further the user is displayed through the recommendation system.
In this context, the module information is composed of operation page-operation module, and field value customization can be performed through a data dictionary, for example: homepage-deposit a0010001, homepage-loan a0010003, credit card-application transaction B0010001, etc.
Control information: the method consists of operation controls, and can carry out field value customization through a data dictionary, such as: 01-inquiry deposit, 02-large deposit, 03-regular deposit, etc.
Operation information: the method consists of operation types, and can carry out field value customization through a data dictionary, such as: 01-inquiry, 02-deposit, 03-withdrawal, 04-update, etc.
Operating time: the format is yyyy-MM-dd hh, MM: ss format collection, and subsequent data processing is convenient.
As shown in fig. 3, the recommendation system includes several modules (each module corresponds to a touch area on the terminal and may be referred to as a module area in this document), and each operation module is further provided with a corresponding operation (each operation corresponds to a touch area on the terminal and may be referred to as an operation area in this document), which illustrates that, since the recommendation system in this document may be applicable to a financial field, the modules may include financial products such as stocks, funds, futures, securities, etc., and the operations may be articles introducing stocks to be good or empty, or may be entries guiding a user to conduct a transaction, or may be an interpretation of a policy by a certain expert, or may have other operation areas, which are not limited in this document.
Herein, the user information may include operator information, region information, website information, system version information, and the like, and the history behavior data includes the number of clicks and stay time of the user for each area within each period of time, and the like.
Herein, the sampling period includes hourly, daily and weekly, but also monthly and yearly, and the time span of the sampling period can be adjusted as required by the person skilled in the art.
In this context, each page in the recommendation system, or each screen in the recommendation system in the terminal, may present several presentation contents (functional areas, herein functional areas include a module area and an operation area).
As one embodiment herein, the method for obtaining the user information and the historical behavior data includes:
embedding a buried point component in the recommendation system, and acquiring the user information and the historical behavior data of the target user to a buried point server through the buried point component;
specifically, the embedded point sdk is added to the recommendation system, and the address of the embedded point server is configured in sdk, and the recommendation system sends the request to the embedded point server through http/https. After the configuration is completed, the embedded point sdk is initialized after the user logs in the recommendation system, and the operation behavior of the user is automatically captured.
When a user starts the recommendation system, the recommendation system sends a request message between acquiring display contents. The request message of the recommendation system contains system related data: user information (operator, region, website, etc.), system information (browser-related data, system version information, etc.), user operation information (operation page information, operation button information, operation start and end times, etc.). After the information is sent to the background embedded point server through the request message, the embedded point server records the information in a log file of a server log.
In the step, the embedded point server and the service server responding to the recommended result of the user are different servers, so that the embedded point data collecting process does not influence the service and cannot cause synchronous blockage of the request and the data.
Transmitting the user information and the historical behavior data to a database according to the proxy file; the agent file comprises Kafka information and buried point field information; the Kafka information is used for determining the address, the login account and the login password of the database; the embedded point field information is used for filtering the user information and data fields in the historical behavior data.
Specifically, the kafka synchronization setting is configured: configuring a proxy file in a buried point server: including Kafka information and buried word segment information. Wherein the Kafka information comprises Kafka addresses, account numbers, passwords and the like; when the embedded field information is message data, the data needing to be input into the database can be pulled according to the proxy file field.
Setting a database warehouse-in algorithm: an automatic binning algorithm is set up to actively pull the kafka message every 15 minutes to half an hour.
In the step, the effect of recording the embedded point data can be achieved only by configuring the address of the embedded point server, the Kafka synchronization related configuration and the database entry table, and the time and labor cost are reduced.
By the method, all behavior data of the user can be saved, so that model training, recommendation prediction and other operations can be performed later.
In the prior art, the recommendation method is not connected with the periodic operation of the user and the periodicity of the financial products, and ignores the periodicity of the financial products. Aiming at the problem, the method records and processes the data by using the time sequence characteristic of the buried point data, predicts the buried point data by establishing a sequence model, can effectively reject misoperation data, retain effective data and predict page contents or modules preferred by users in a future period of time. The method establishes a model, analyzes and predicts the user behavior, and puts preferred marketing content into the user according to the personalized result obtained by prediction, thereby reducing the risk of putting advertisement errors and improving the user experience.
In order to enrich the recommended content in the recommendation system, a plurality of different prediction models can be set for the user, and each prediction model is used for respectively predicting a plurality of contents to be displayed, fusing all the contents to be displayed and displaying the contents to the user. Therefore, the effect of enriching the recommended content is achieved, and the behavior of the user can be more fitted.
As one embodiment herein, the sampling period comprises hourly, daily, or weekly; the type of the prediction model corresponds to the sampling period type, and the prediction model comprises a first prediction model, a second prediction model and a third prediction model;
the step of importing the average residence time of each functional area into the prediction model to obtain the content to be displayed of the target user in the recommendation system, further comprises the steps of:
the average residence time of each functional area calculated according to each hour is imported into a first prediction model to obtain a first sub-set to be displayed of the target user in the recommendation system;
the average residence time of each functional area obtained through daily calculation is imported into a second prediction model, and a second sub-set to be displayed of the target user in the recommendation system is obtained;
The average residence time of each functional area obtained through weekly calculation is imported into a third prediction model, and a third sub-set to be displayed of the target user in the recommendation system is obtained;
and fusing the first sub-set to be displayed, the second sub-set to be displayed and the third sub-set to be displayed to obtain the content to be displayed of the target user in the recommendation system.
In this context, for the same historical behavior data, different data may be obtained after processing using different sampling periods, i.e. the average residence time of a functional area is calculated using different sampling periods, and the final average residence time may be different. Different predictive models, namely a first predictive model, a second predictive model and a third predictive model, can be obtained after training the same predictive model with different data.
As an embodiment of the present disclosure, the obtaining, according to a sampling period and the historical behavior data, an average residence time of the target user in the sampling period for each functional area of the recommendation system further includes:
determining the clicking times of the target user for each functional area in the recommendation system according to the historical behavior data in the sampling period;
Determining the stay time of the target user in each functional area according to the time interval of the two functional areas continuously clicked by the target user;
and in the sampling period, determining the average residence time of the functional areas in the sampling period according to the total residence time of each functional area and the clicking times of each functional area.
In this context, if the sampling period is one day, the number of clicks of the user on a certain functional area in one day may be counted, if the functional area is a stock, the number of clicks of the user on the stock in one day may be counted, and if the functional area is an a operation in the stock module, the number of clicks of the a operation may be counted. In this way, the number of clicks of all the functional areas is obtained.
Then by the time interval of two functional areas of two consecutive clicks by the user, e.g. 16:30 clicks on the A operation and 17:00 clicks on the B operation, then the user may be considered to have remained on the A operation for 30 minutes.
Calculating the single stay time t of each functional area according to the following formula ai ：
t ai ＝x 1 -ai 1
Wherein t is ai Representing the single dwell time length x of the target user in the ith dwell function area a in the sampling period 1 Indicating the time ai for recording the buried data when the functional area a jumps to the next functional area 1 The first buried data recording time of the ith stay in the functional area a in a certain hour is represented;
in this context, the term "a" is used herein,
calculating the total stay time T of the target user in the functional area in the sampling period according to the following formula a
Wherein T is a T is the total residence time in the sampling period ai Representing the single stay time length of the target user staying in the functional area for the ith time in the sampling period, wherein i is the ith time, and n is the total stay times in the sampling period;
calculating the average residence time v of the target user in the functional area in the sampling period according to the following formula a
Wherein n is a T is the total number of clicks of the functional area a V is the total dwell time of the functional area in the sampling period a For the average residence time of the target user in the functional area during the sampling period.
For example, by the above formula, the average residence time resulting in different sampling periods is different in the same behavior data.
Taking the example in hours: and acquiring all user behavior data of a certain user in a week, and carrying out statistics and processing on the behavior data according to the hours.
Module information, control information and operation type information of each hour in a week are counted: frequency calculation is performed on the above three fields, such as: the number of clicks of "homepage-deposit a0010001", the number of clicks of "01-inquiry deposit", and the number of clicks of "01-inquiry" for a certain hour in the week. By means of the statistics, the number of button clicks of the module and the button operated by each module can be counted every hour in the week of the user.
By acquiring different module data, the dwell time of the previous module can be acquired. First, two adjacent different pieces of module data a are positioned n And b 1 Then locating a piece of data a which appears earliest in the module in the previous piece of data 1 The piece of data and a is then calculated 1 And b 1 The time difference between them, the time the user stays in this module is obtained. Such as: a user operates "01 inquiry" - "01 inquiry deposit", "02-deposit" - "02-large deposit", "04-update" - "03-regular deposit" (one operation is one piece of buried point data, which is three pieces of buried point data) on "homepage-deposit a0010001" for a certain hour, and then jumps to the "credit card-application card B0010001" module, and then the user stays at "homepage-deposit a0010001" for a time between the first piece of buried point data appearing in "homepage-deposit a0010001" and the jump to "credit card-application card B0010001", respectively.
Knowing the total residence time of each functional area and the number of clicks of each functional area in each hour, and calculating the average residence time v of the target user in the functional area in the sampling period according to the following formula a
Wherein n is a T is the total number of clicks of the functional area a For the total dwell time of the functional area within the sampling periodLong, v a For the average residence time of the target user in the functional area during the sampling period.
As an embodiment herein, the fusing the first sub-to-be-displayed set, the second sub-to-be-displayed set, and the third sub-to-be-displayed set to obtain to-be-displayed content of the target user in the recommendation system further includes:
determining whether the same display content exists in the first sub-set to be displayed, the second sub-set to be displayed and the third sub-set to be displayed, and if so, displaying the display content in a primary area in the recommendation system;
determining whether the same display content exists in any two sub-to-be-displayed sets of the first sub-to-be-displayed set, the second sub-to-be-displayed set and the third sub-to-be-displayed set, and if so, displaying the display content in a secondary area in the recommendation system.
Herein, the primary region may be a user-defined region, i.e., a region most frequently used according to a user's click habit, such as a middle position of a terminal, and the secondary region may be a user-defined region, i.e., a region less frequently used according to a user's click habit, such as a middle left or middle right position of a terminal. The specific text is not limited.
By the method, pushing can be more accurate, a user can conveniently watch the content which the user wants to see, and the pushing mode is more accurate.
In order to better understand the modeling concept of the predictive model herein, a specific training process of the predictive model is given herein, and before training of the predictive model begins, basic filtering of the data is required to obtain stable data. Before the sequence model is built, the stationarity of the data needs to be checked. First, the distribution of data needs to be observed. If the data fluctuates up and down around 0 as a whole or floats up and down on a certain horizontal line, it can be preliminarily judged as stable data.
And positioning the head-tail abnormal data in the historical behavior data according to t distribution in a t test method, and removing the head-tail abnormal data in the historical behavior data according to a stability test result to obtain stable data.
ADF detection (Augmented Dickey-Fuller test), also known as the Augmented Diety-Fowler test, may also be used to determine whether behavior data is stationary using the autoregressive equation in its detection method:
Y t ＝BY t-1 +A+∈ t
in autoregressive, if the lag term coefficient B is 1, it can be regarded that there is a unit root, namely, colloquially called random walk (random walk).
As an embodiment herein, the prediction model training method shown in fig. 4 includes:
step 401, dividing historical behavior data in different sampling periods into a training set and a testing set in proportion;
in this step, the historical behavior data processed in the database is acquired, and different sampling periods are used for processing the same batch of historical behavior data. In this context, a sampling period may include hourly, daily, and weekly, and then three types of processed historical behavioral data may be obtained for the same historical behavioral data.
And dividing each processed historical behavior data into a training set and a testing set. The first kind of history behavior data obtained by the first processing is divided into training set and testing set. The historical behavior data of the second processing completion obtained finally is divided into a training set and a testing set by taking each day as a sampling period. The historical behavior data of the third processing completion obtained finally is divided into a training set and a testing set by taking each week as a sampling period.
Step 402, respectively importing the training set into a random sequence model and a periodic sequence model, and performing repeated iterative training to obtain a random sequence model and a periodic sequence model, wherein the random sequence model and the periodic sequence model are input by corresponding numbers of all functional areas and average residence time of all functional areas, and the corresponding numbers of all functional areas at the next moment are output;
In this step, since the fitting effect of different models on the data is different, two models are used herein to train using training sets. The random sequence model is in the form of:
Y t ＝(1+β 1 )Y t-1 +···+(β p -β p-1 )Y t-p +e t -θ 1 e t-1 -θ 2 e t-2 -···-θ q e t-q
the characteristic equation is:
wherein B represents profile expression over time, beta (B) represents AR (p) model, theta (B) represents MA (q) model, Y t For output at time t, Y t-1 For output at time t-1, Y t-p For output at time t-p, beta 1 、β p 、β p-1 、θ 1 、θ 2 、θ q Representing weight parameters e t 、e t-1 、e t-2 、e t-q Is a plateau parameter.
Wherein the periodic sequence model is in the form of:
Φ(B s )＝1-ΦB 2s -ΦB 2s -…-Φ p B ps
Θ(B s )＝1-ΘB 2s -ΘB 2s -…-Θ p B Qs
combining the two formulas to obtain
Finally, the periodic sequence model is expressed as
ARIMA(p,d,q)×(P,D,Q)
Wherein P is the order of seasonal autoregressive, D is the number of times of seasonal difference, Q is the order of seasonal moving average, B is a delay operator, phi (B) is a delay operator polynomial, s is the length of the season, P is the order of non-seasonal autoregressive, D is the number of times of one-step difference, Q is the order of non-seasonal moving average, and theta (B) is calculated according to phi (B).
The periodic sequence model is suitable for the data in a periodic form, and the model can better obtain the data possibly appearing at the next moment aiming at the data in the periodic form.
In this context, each functional area has its unique organization, e.g., a user operating "01 inquiry" - "01 inquiry deposit", "02-deposit" - "02-bulk deposit", "04-update" - "03-regular deposit" (an operation as a piece of buried point data, three pieces of buried point data) at "home-deposit A0010001" for a certain hour, respectively, and then jumping to the "credit card-application transaction B0010001" module.
In brief, the input in the sequence model may comprise two parts, the first part being the number of the different functional modules and the second part being the average residence time of the different functional modules. The output of the sequence model may be the code corresponding to the functional module that the user may click on at the next time. For example, 01 inquiry "-"01 inquiry deposit "," 02-deposit "-" 02-large deposit "," 04-update "-" 03-regular deposit, etc.
And step 403, testing the trained random sequence model and the trained periodic sequence model by using a test set, and taking the sequence model with the smallest difference between the test result and the test set as a prediction model.
In this context, the same number of iterations is used to derive two sequence models for each sample period of behavior data. And then using the test set to verify the two sequence models, and finally taking the sequence model with the best fitting effect as a prediction model. In this context, the difference method may be used to calculate the difference between the test result and the test set, e.g. the true value in the test set is 100 and the test result is 80, then the difference is 20.
A message pushing device as shown in fig. 5, comprising:
An obtaining unit 501, configured to obtain user information of a target user in response to the target user logging into the recommendation system;
an extracting unit 502, configured to extract, according to the user information, a trained prediction model corresponding to the target user, and historical behavior data in a sampling period; the historical behavior data comprise operation module data, operation button data, operation starting time and operation ending time;
an importing unit 503, configured to obtain, according to a sampling period and the historical behavior data, an average residence time of the target user in the sampling period for each functional area of the recommendation system, where the functional area includes a module area and an operation area;
and the prediction unit 504 is configured to import the average residence time of each functional area into the prediction model, so as to obtain content to be displayed of the target user in the recommendation system, where the display content includes each functional area in the recommendation system.
By adopting the technical scheme, the user information of the target user is acquired by responding to the login recommendation system of the target user, so that the user information corresponding to the identity of the target user can be acquired; extracting a prediction model which corresponds to the target user and is trained according to the user information, and historical behavior data in a sampling period; the historical behavior data comprise operation module data, operation button data, operation starting time and operation ending time, and a prediction model corresponding to a target user and past historical behavior data can be obtained; the average stay time of the target user in the sampling period for each functional area of the recommendation system is obtained according to the sampling period and the historical behavior data, wherein the functional area comprises a module area and an operation area, so that the stay time of the target user in a plurality of sampling periods for each operation of different functional areas can be determined, and the average stay time of the target user in the sampling period is further determined through the stay time of each operation; the average residence time of each functional area is imported into the prediction model to obtain the content to be displayed of the target user in the recommendation system, wherein the display content comprises each functional area in the recommendation system, the content to be displayed which is relatively close to the content actually required by the target user can be obtained, and further the user is displayed through the recommendation system.
As shown in fig. 6, for one computer device provided by embodiments herein, the computer device 602 may include one or more processors 604, such as one or more Central Processing Units (CPUs), each of which may implement one or more hardware threads. The computer device 602 may also include any memory 606 for storing any kind of information, such as code, settings, data, etc. For example, and without limitation, memory 606 may include any one or more of the following combinations: any type of RAM, any type of ROM, flash memory devices, hard disks, optical disks, etc. More generally, any memory may store information using any technique. Further, any memory may provide volatile or non-volatile retention of information. Further, any memory may represent fixed or removable components of computer device 602. In one case, when the processor 604 executes associated instructions stored in any memory or combination of memories, the computer device 602 can perform any of the operations of the associated instructions. The computer device 602 also includes one or more drive mechanisms 608, such as a hard disk drive mechanism, an optical disk drive mechanism, and the like, for interacting with any memory.
The computer device 602 may also include an input/output module 610 (I/O) for receiving various inputs (via an input device 612) and for providing various outputs (via an output device 614). One particular output mechanism may include a presentation device 616 and an associated Graphical User Interface (GUI) 618. In other embodiments, input/output module 610 (I/O), input device 612, and output device 614 may not be included, but may be implemented as a single computer device in a network. The computer device 602 may also include one or more network interfaces 620 for exchanging data with other devices via one or more communication links 622. One or more communication buses 624 couple the above-described components together.
The communication link 622 may be implemented in any manner, for example, through a local area network, a wide area network (e.g., the internet), a point-to-point connection, etc., or any combination thereof. Communication link 622 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc., governed by any protocol or combination of protocols.
Corresponding to the method in fig. 2-4, embodiments herein also provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the above method.
Embodiments herein also provide a computer readable instruction wherein the program therein causes the processor to perform the method as shown in fig. 2-4 when the processor executes the instruction.
It should be understood that, in the various embodiments herein, the sequence number of each process described above does not mean the sequence of execution, and the execution sequence of each process should be determined by its functions and internal logic, and should not constitute any limitation on the implementation process of the embodiments herein.
It should also be understood that in embodiments herein, the term "and/or" is merely one relationship that describes an associated object, meaning that three relationships may exist. For example, a and/or B may represent: a exists alone, A and B exist together, and B exists alone. In addition, the character "/" herein generally indicates that the front and rear associated objects are an "or" relationship.
Those of ordinary skill in the art will appreciate that the elements and algorithm steps described in connection with the embodiments disclosed herein may be embodied in electronic hardware, in computer software, or in a combination of the two, and that the elements and steps of the examples have been generally described in terms of function in the foregoing description to clearly illustrate the 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 solution. 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 disclosure.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, and are not repeated herein.
In the several embodiments provided herein, it should be understood that the disclosed systems, devices, and methods may be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. In addition, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices, or elements, or may be an electrical, mechanical, or other form of connection.
The units described as separate units may or may not be physically separate, and units shown 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 elements may be selected according to actual needs to achieve the objectives of the embodiments herein.
In addition, each functional unit in the embodiments herein may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solutions herein are essentially or portions contributing to the prior art, or all or portions of the technical solutions may be embodied in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments herein. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Specific examples are set forth herein to illustrate the principles and embodiments herein and are merely illustrative of the methods herein and their core ideas; also, as will be apparent to those of ordinary skill in the art in light of the teachings herein, many variations are possible in the specific embodiments and in the scope of use, and nothing in this specification should be construed as a limitation on the invention.
Claims (11)
1. A message pushing method, comprising:
responding to the login of a target user to a recommendation system, and acquiring user information of the target user;
extracting a prediction model which is trained and completed and corresponds to the target user and historical behavior data in a sampling period according to the user information; the historical behavior data comprise operation module data, operation button data, operation starting time and operation ending time;
obtaining average residence time of the target user in the sampling period for each functional area of the recommendation system according to the historical behavior data, wherein the functional areas comprise a module area and an operation area;
and importing the corresponding numbers of the functional areas and the average residence time of the functional areas into the prediction model to obtain the content to be displayed of the target user in the recommendation system, wherein the display content comprises the functional areas in the recommendation system, and the prediction model is trained by using the historical behavior data.
2. The message pushing method of claim 1, wherein the sampling period comprises hourly, daily, or weekly; the prediction model comprises a first prediction model, a second prediction model and a third prediction model;
the step of importing the average residence time of each functional area into the prediction model to obtain the content to be displayed of the target user in the recommendation system, further comprises the steps of:
the average residence time of each functional area calculated according to each hour is imported into a first prediction model to obtain a first sub-set to be displayed of the target user in the recommendation system;
the average residence time of each functional area obtained through daily calculation is imported into a second prediction model, and a second sub-set to be displayed of the target user in the recommendation system is obtained;
the average residence time of each functional area obtained through weekly calculation is imported into a third prediction model, and a third sub-set to be displayed of the target user in the recommendation system is obtained;
and fusing the first sub-set to be displayed, the second sub-set to be displayed and the third sub-set to be displayed to obtain the content to be displayed of the target user in the recommendation system.
3. The message pushing method according to claim 2, wherein the fusing the first sub-to-be-displayed set, the second sub-to-be-displayed set, and the third sub-to-be-displayed set to obtain to-be-displayed content of the target user in the recommendation system, further includes:
determining whether the same display content exists in the first sub-set to be displayed, the second sub-set to be displayed and the third sub-set to be displayed, and if so, displaying the display content in a primary area in the recommendation system;
determining whether the same display content exists in any two sub-to-be-displayed sets of the first sub-to-be-displayed set, the second sub-to-be-displayed set and the third sub-to-be-displayed set, and if so, displaying the display content in a secondary area in the recommendation system.
4. The message pushing method according to claim 1, wherein the obtaining method of the user information and the historical behavior data includes:
acquiring the user information and the historical behavior data of the target user to a buried point server through a buried point component which is embedded in advance;
Transmitting the user information and the historical behavior data in the embedded point server to a database according to a proxy file; the agent file comprises Kafka information and buried point field information; the Kafka information is used for determining the address, the login account and the login password of the database; the embedded point field information is used for filtering the user information and data fields in the historical behavior data.
5. The message pushing method according to claim 1, wherein the obtaining, according to the historical behavior data, an average residence time of the target user in the sampling period for each functional area of the recommendation system further includes:
determining the clicking times of the target user for each functional area in the recommendation system according to the historical behavior data in the sampling period;
determining the time length of each stay of the target user in each functional area according to the time interval of the two functional areas continuously clicked by the target user;
and in the sampling period, determining the average residence time of the functional areas in the sampling period according to the residence time of each functional area and the clicking times of each functional area.
6. The message pushing method according to claim 5, wherein determining the duration of stay of the target user in each functional area according to the time interval of two functional areas continuously clicked by the target user comprises: calculating the single stay time t of each functional area according to the following formula ai ：
t ai ＝x 1 -ai 1
Wherein t is ai Representing the single dwell time length x of the target user in the ith dwell function area a in the sampling period 1 Indicating the time ai for recording the buried data when the functional area a jumps to the next functional area 1 The first buried data recording time of the ith stay in the functional area a in a certain hour is represented;
and in the sampling period, determining the average residence time of the functional areas in the sampling period according to the residence time of each functional area and the clicking times of each functional area, and further comprising:
calculating the total stay time T of the target user in the functional area in the sampling period according to the following formula a
Wherein T is a T is the total residence time in the sampling period ai Representing the single stay time length of the target user staying in the functional area for the ith time in the sampling period, wherein i is the ith time, and n is the total stay times in the sampling period;
Calculating the sampling period according to the following formulaAverage residence time v of target user in functional area a
Wherein n is a T is the total number of clicks of the functional area a V is the total dwell time of the functional area in the sampling period a For the average residence time of the target user in the functional area during the sampling period.
7. The message pushing method of claim 5, wherein the predictive model training method comprises:
dividing historical behavior data in different sampling periods into a training set and a testing set in proportion;
respectively importing the training set into a random sequence model and a periodic sequence model, and performing repeated iterative training to obtain a random sequence model and a periodic sequence model, wherein the random sequence model and the periodic sequence model are input by corresponding numbers of all functional areas and average residence time of all the functional areas, and the random sequence model and the periodic sequence model are output by corresponding numbers of all the functional areas at the next moment;
and testing the trained random sequence model and the trained periodic sequence model by using a test set, and taking the sequence model with the smallest difference between the test result and the test set as a prediction model.
8. The message pushing method of claim 7, comprising, prior to scaling historical behavior data over different sampling periods into a training set and a test set:
And positioning the head-tail abnormal data in the historical behavior data according to t distribution in a t test method, and removing the head-tail abnormal data in the historical behavior data according to a stability test result.
9. A message pushing device, comprising:
the acquisition unit is used for responding to the login of the target user to the recommendation system and acquiring the user information of the target user;
the extraction unit is used for extracting a prediction model which corresponds to the target user and is trained according to the user information and historical behavior data in a sampling period; the historical behavior data comprise operation module data, operation button data, operation starting time and operation ending time;
the importing unit is used for obtaining average stay time of the target user for each functional area of the recommendation system in the sampling period according to the historical behavior data, wherein the functional areas comprise a module area and an operation area;
the prediction unit is used for importing the corresponding numbers of the functional areas and the average residence time of the functional areas into the prediction model to obtain the content to be displayed of the target user in the recommendation system, wherein the display content comprises the functional areas in the recommendation system, and the prediction model is trained by using the historical behavior data.
10. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the message pushing method according to any of claims 1-8 when executing the computer program.
11. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program which, when executed by a processor, implements the message pushing method according to any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311609910.5A CN117421485A (en) | 2023-11-29 | 2023-11-29 | Message pushing method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311609910.5A CN117421485A (en) | 2023-11-29 | 2023-11-29 | Message pushing method, device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117421485A true CN117421485A (en) | 2024-01-19 |
Family
ID=89528474
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311609910.5A Pending CN117421485A (en) | 2023-11-29 | 2023-11-29 | Message pushing method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117421485A (en) |
-
2023
- 2023-11-29 CN CN202311609910.5A patent/CN117421485A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101708444B1 (en) | Method for evaluating relation between keyword and asset value and Apparatus thereof | |
| CN111898023A (en) | Message pushing method and device, readable storage medium and computing equipment | |
| CN107784504B (en) | Method for generating return visit event of client and terminal equipment | |
| CN108520440B (en) | Method for determining total consumption stock demand of reading user, electronic equipment and computer storage medium | |
| US20120311140A1 (en) | Method of processing web access information and server implementing same | |
| CN111028087A (en) | Information display method, device and equipment | |
| US11410203B1 (en) | Optimized management of online advertising auctions | |
| CN117235586B (en) | Hotel customer portrait construction method, system, electronic equipment and storage medium | |
| CN112561565A (en) | User demand identification method based on behavior log | |
| US20140082480A1 (en) | Identification of sequential browsing operations | |
| CN108780446A (en) | The prompt that time dependent machine generates | |
| CN111723083A (en) | User identity identification method and device, electronic equipment and storage medium | |
| US20240095766A1 (en) | Merchant Advertisement Informed Item Level Data Predictions | |
| CN102769625A (en) | Client-side Cookie information acquisition method and device | |
| CN114385942A (en) | Page element parameter optimization method, device and equipment based on user behavior | |
| US10776138B1 (en) | Recipient-side dynamic customization of an email intended for the recipient | |
| EP3082090A1 (en) | Method for determining a user profile in relation to certain web content | |
| CN117421485A (en) | Message pushing method, device and storage medium | |
| CN114663139A (en) | Resource bit processing method and device, storage medium and processor | |
| CN111026981B (en) | Visual display method, device and equipment for hot topics | |
| CN113836430A (en) | Book recommendation method, terminal and storage medium | |
| CN110727797A (en) | Label generation method and device, electronic equipment and computer readable medium | |
| US20230350969A1 (en) | Automatic display control method for web content | |
| CN112243156B (en) | Barrage display method and device and storage medium | |
| CN111881355B (en) | Object recommendation method and device, storage medium and processor |
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 |