CN113487389A - Information recommendation method and device - Google Patents
Information recommendation method and device Download PDFInfo
- Publication number
- CN113487389A CN113487389A CN202110850844.5A CN202110850844A CN113487389A CN 113487389 A CN113487389 A CN 113487389A CN 202110850844 A CN202110850844 A CN 202110850844A CN 113487389 A CN113487389 A CN 113487389A
- Authority
- CN
- China
- Prior art keywords
- user
- target
- model
- characteristic data
- hierarchy
- 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 67
- 238000012549 training Methods 0.000 claims description 29
- 238000012545 processing Methods 0.000 claims description 25
- 238000004140 cleaning Methods 0.000 claims description 10
- 238000009795 derivation Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 description 14
- 238000000354 decomposition reaction Methods 0.000 description 9
- 238000001914 filtration Methods 0.000 description 7
- 238000013528 artificial neural network Methods 0.000 description 5
- 238000012795 verification Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013486 operation strategy Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- 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
- 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
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0277—Online advertisement
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application discloses an information recommendation method and device, and the method comprises the following steps: and acquiring user characteristic data of the target user. And inputting the user characteristic data into the user model to obtain user images of the target user in different scenes. Based on the user image of the target user in the target scene, the user hierarchy of the target user in the target scene may be determined. Based on the user hierarchy of the target user in the target scene, a target recommendation model corresponding to the target scene can be determined. And outputting the target product in the target scene by using the user hierarchy of the target user in the target scene and the target recommendation model. And recommending the target product information in the target scene to the target user. The method achieves the purposes of acquiring the target product information under different scenes according to the user characteristics of the target user and automatically recommending the target product information to the target user.
Description
Technical Field
The application relates to the technical field of internet, in particular to an information recommendation method and device.
Background
Generally, student users perform different activities in different scenes in a campus, and different data is generated. For example, when a student user borrows a book from a library in a campus, borrowing data can be generated; the student user purchases the product and generates payment data.
Based on this, it is very necessary to carry out the accurate recommendation of different product information to student's user in different campus scenes, and the demand of student's user can be satisfied through automatic recommendation product information.
Disclosure of Invention
In order to solve the technical problem, the application provides an information recommendation method and device, which are used for recommending product information to a user accurately according to different scenes so as to meet the user requirements.
In order to achieve the above purpose, the technical solutions provided in the embodiments of the present application are as follows:
the embodiment of the application provides an information recommendation method, which comprises the following steps:
acquiring user characteristic data of a target user;
inputting the user characteristic data into a user model, and acquiring user images of the target user in different scenes; the user model is used for acquiring user images of the target user in different scenes according to the user characteristic data of the target user;
acquiring user hierarchy of the target user in the target scene according to the user image of the target user in the target scene; the target scene is any one of the different scenes;
determining a target recommendation model corresponding to the target scene according to the target scene corresponding to the target user;
acquiring a target product of a target scene by using the user hierarchy of the target user in the target scene and the target recommendation model, and recommending product information corresponding to the target product to the target user; the target recommendation model is used for outputting target products corresponding to the target scene according to user hierarchy of the target scene.
Optionally, the training process of the user model includes:
acquiring historical user characteristic data and a label corresponding to the historical user characteristic data;
and training a user model according to the historical user characteristic data and the label corresponding to the historical user characteristic data until a preset condition is reached, and acquiring the trained user model.
Optionally, before the obtaining of the user characteristic data of the target user, the method further includes:
acquiring original data of a target user;
and carrying out data cleaning processing and data derivation processing on the original data of the target user to generate user characteristic data of the target user.
Optionally, when the target scenes are different, the model parameters of the target recommendation model are different.
Optionally, the user hierarchy includes one or more of a demographic dimension hierarchy, a credit dimension hierarchy, a consumption dimension hierarchy, and an hobby dimension hierarchy.
An embodiment of the present application further provides an information recommendation device, where the device includes:
the first acquisition unit is used for acquiring user characteristic data of a target user;
the input unit is used for inputting the user characteristic data into a user model and acquiring user images of the target user in different scenes; the user model is used for acquiring user images of the target user in different scenes according to the user characteristic data of the target user;
the second acquisition unit is used for acquiring the user hierarchy of the target user in the target scene according to the user image of the target user in the target scene; the target scene is any one of the different scenes;
the determining unit is used for determining a target recommendation model corresponding to the target scene according to the target scene corresponding to the target user;
the recommendation unit is used for acquiring a target product of a target scene by using the user hierarchy of the target user in the target scene and the target recommendation model, and recommending product information corresponding to the target product to the target user; the target recommendation model is used for outputting target products corresponding to the target scene according to user hierarchy of the target scene.
Optionally, the apparatus further comprises: a model training unit;
the model training unit comprises:
the acquisition subunit is used for acquiring historical user characteristic data and a label corresponding to the historical user characteristic data;
and the training subunit is used for training the user model according to the historical user characteristic data and the labels corresponding to the historical user characteristic data until a preset condition is reached, and acquiring the trained user model.
Optionally, the apparatus further comprises:
a third obtaining unit, configured to obtain original data of a target user before obtaining user feature data of the target user;
and the processing unit is used for carrying out data cleaning processing and data derivation processing on the original data of the target user to generate user characteristic data of the target user.
Optionally, the user hierarchy includes one or more of a demographic dimension hierarchy, a credit dimension hierarchy, a consumption dimension hierarchy, and an hobby dimension hierarchy.
The embodiment of the application also provides an information recommendation system, which at least comprises the campus multi-scene one-card service platform, and the campus multi-scene one-card service platform executes the information recommendation method.
According to the technical scheme, the method has the following beneficial effects:
the embodiment of the application provides an information recommendation method and device, and user characteristic data of a target user are obtained. And inputting the user characteristic data into the user model to obtain user images of the target user in different scenes. Based on the user image of the target user in the target scene, the user hierarchy of the target user in the target scene may be determined. Based on the user hierarchy of the target user in the target scene, a target recommendation model corresponding to the target scene can be determined. And outputting the target product in the target scene by using the user hierarchy of the target user in the target scene and the target recommendation model. And recommending the target product information in the target scene to the target user. The method achieves the purposes of acquiring the target product information under different scenes according to the user characteristics of the target user and automatically recommending the target product information to the target user.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic diagram of an exemplary application scenario of an information recommendation method provided in an embodiment of the present application;
fig. 2 is a flowchart of an information recommendation method according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of an information recommendation device according to an embodiment of the present application.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, embodiments accompanying the drawings are described in detail below.
In order to facilitate understanding and explaining the technical solutions provided by the embodiments of the present application, the following first describes the background art of the embodiments of the present application.
In recent years, payment services for student users in campuses have been rapidly developed. When the student users consume in different scenes in the campus, some products and services which meet the interests of the student users are expected to be recommended so as to meet the requirements of purchasing the products or services.
Based on this, the embodiment of the application provides an information recommendation method and device, and user characteristic data of a target user is obtained. The user characteristic data may represent the target user's own characteristics and payment characteristics. And inputting the user characteristic data into the user model to obtain user images of the target user in different scenes. Based on the user image of the target user in the target scene, the user hierarchy of the target user in the target scene may be determined. Based on the target scene corresponding to the target user, the target recommendation model corresponding to the target scene can be determined. And outputting the target product in the target scene by using the user hierarchy of the target user in the target scene and the target recommendation model. And recommending the target product information in the target scene to the target user. The method achieves the purposes of acquiring the target product information under different scenes according to the user characteristics of the target user and automatically recommending the target product information to the target user.
In order to facilitate understanding of the information recommendation method provided in the embodiment of the present application, the following description is made with reference to a scenario example shown in fig. 1. Referring to fig. 1, the figure is a schematic view of an exemplary application scenario of an information recommendation method provided in an embodiment of the present application. The method is applicable to the terminal 101.
And binding the all-purpose card and starting to use by the target user, wherein the target user is any student user. The data generated by the target user through using the one-card is stored on the campus multi-scene one-card service platform 102. The terminal 101 acquires user feature data of a target user through the campus multi-scenario one-card service platform 102. The user characteristic data of the target user is data generated by the target user through using the one-card.
After the terminal 101 obtains the user feature data, the user feature data needs to be input into the user model, and user images of the target user in different use scenes are obtained. And determining the user layering corresponding to the user image of the target scene according to the user image of the target scene. The target scene is any one of different usage scenes.
The terminal 101 determines a target recommendation model corresponding to the target scene according to the target scene. And acquiring a target product of the target scene according to the user hierarchy of the target user in the target scene and the target propulsion model, and recommending the target product information to the target user in the target scene.
Those skilled in the art will appreciate that the block diagram shown in fig. 1 is only one example in which embodiments of the present application may be implemented. The scope of applicability of the embodiments of the present application is not limited in any way by this framework.
It is noted that terminal 101 can be any user equipment now known, developing or later developed that is capable of interacting with each other through any form of wired and/or wireless connection (e.g., Wi-Fi, LAN, cellular, coaxial cable, etc.), including but not limited to: smart wearable devices, smart phones, non-smart phones, tablets, laptop personal computers, desktop personal computers, minicomputers, midrange computers, mainframe computers, and the like, either now in existence, under development, or developed in the future. The embodiments of the present application are not limited in any way in this respect. It should also be noted that the server 102 in the embodiment of the present application may be an example of an existing device, a device under development, or a device developed in the future, which is capable of performing the above operations. The embodiments of the present application are not limited in any way in this respect.
Based on the above description, the information recommendation method provided in the present application will be described in detail below with reference to the accompanying drawings.
Referring to fig. 2, fig. 2 is a flowchart of an information recommendation method according to an embodiment of the present application. As shown in fig. 2, the information recommendation method includes S201 to S205:
s201: and acquiring user characteristic data of the target user.
The target user is specifically any one of the student users. Student users bind the one-card and use the one-card, and data generated by using the one-card is stored in the campus multi-scene one-card service platform.
As an example, the campus multi-scenario one-card service platform can be arranged in a bank application program, and the bank application program is located in a terminal. The method provided by the embodiment of the application is executed by a campus multi-scene one-card service platform. The campus multi-scene one-card service platform directly obtains user feature data of a target user, analyzes a target product conforming to the target user according to the user feature data, and recommends target product information to the target user. The target products comprise learning-aid resources, financial services and the like.
As another example, the method provided by the embodiment of the present application is executed by a terminal, and the terminal acquires user feature data of a target user from a campus multi-scenario one-card service platform.
The user characteristic data at least comprises one or more of user gender, user age, user region, user behavior, user interest, user use duration, user industry field, user login times and user consumption capacity. The method provided by the embodiment of the application aims to acquire the target product information recommended to the target user by using the user characteristic data of the target user.
It should be noted that before obtaining the user feature data of the target user, obtaining that the target user is raw data, and processing the raw data to obtain the user feature data of the target user, please refer to the following specific processing procedure.
S202: inputting the user characteristic data into a user model, and acquiring user images of a target user in different scenes; the user model is used for acquiring user images of the target user in different scenes according to the user characteristic data of the target user.
After the user characteristic data of the target user is obtained, the user characteristic data of the target user is input into the user model to be processed, and user images of the target user in different scenes are obtained. The different scenes are, for example, a library scene and various consumption scenes. For example, the various consumption scenes are consumption scenes of eating, wearing, living, going, financial product consumption scenes and the like.
The user portrait is a tool for describing the user, and describes all-round characteristics of individual users or user groups, so that information such as user preference and behavior is provided for operation analysts, operation strategies are optimized, and accurate user role information is provided for products so as to carry out targeted product information recommendation.
The user model is used for inputting user characteristic data of a target user, classifying and combining the user characteristic data according to different scenes, and forming user images under different scenes.
User portrayal in different scenarios may be described in one or more dimensions. For example, the dimensions include at least one or more of a demographic dimension, a credit attribute dimension, a consumption characteristics dimension, and an interest and hobby dimension.
The demographic attributes dimension is used to describe the user's basic characteristic information including name, gender, age group, phone number, mailbox, home address, etc.
The credit attribute dimension is used to describe the user's revenue potential, revenue situation, and payment capabilities, etc., including asset information and credit information. Income, assets, liabilities, academic records, credit scores, etc. all belong to the credit information.
The consumption characteristic dimension is used for describing the main consumption habits, the consumption preferences, the consumption ability and the purchase frequency of the user and the like. The user can be identified as a specific consumption characteristic crowd according to the consumption record of the user, such as a tourist crowd, a catering user, a financial crowd, a researcher and the like.
The interest dimension is used to describe the user's interests. The consumption tendency of the user can be known according to the interest and hobbies of the user. The user may be characterized as a particular hobby feature population based on the user's consumption record. Such as outdoor sports enthusiasts, travel enthusiasts, movie enthusiasts, science and technology enthusiasts, fitness enthusiasts, reading enthusiasts, and the like.
For example, user characteristic data of the target user (for example, the name, age, academic calendar, book borrowing time, book borrowing name, book returning time, and the like of the user) is input into the user model, and the obtained user portrait of the target user in the library scene is information such as the name and age range of the population attribute dimension, information such as credit score and academic calendar of the credit attribute dimension, information such as the number of times of book borrowing of the consumption characteristic dimension, and the interest and preference dimension is information such as the type and preference of the suspicion book.
It will be appreciated that the user representation of the target user is a process of rendering and categorizing the user characteristic data of the target user. For example, when the user feature data is a name, the user figure is reproduced. And when the user characteristic data is that the xx scout novels are borrowed, and the number of times exceeds a borrowing number threshold, outputting the user portrait as the suspicion book type preference in the interest and preference dimension. User portrayal can be thought of as the process of abstracting the concrete information of user feature data into multiple labels. A plurality of tags are used to represent users.
In a possible implementation manner, an embodiment of the present application provides a specific implementation manner of a training process of a user model, which is specifically referred to below.
S203: acquiring user hierarchies of target users in a target scene according to user images of the target users in the target scene; the target scene is any one of different scenes.
After the user portrait of the target user in the target scene is obtained, the user hierarchy of the target user in the target scene is obtained based on the user portrait.
Correspondingly, the user hierarchy corresponding to the target scene comprises one or more of a population dimension hierarchy, a credit dimension hierarchy, a consumption dimension hierarchy and an interest dimension hierarchy. For example, in the consumption dimension hierarchy in any one consumption scenario, a user may be classified into a hierarchy with high consumption capability and high consumption quality, and a hierarchy with low consumption capability and low consumption quality.
It can be understood that the hierarchy of user hierarchies needs to be divided according to the actual scene. It should be noted that the process from user representation to user layering is a more abstract process. According to the user hierarchy corresponding to the target user, target product information recommendation can be more accurately performed on the target user.
S204: and determining a target recommendation model corresponding to the target scene according to the target scene corresponding to the target user.
When the target scenes are different, the target recommendation models are different, and the model parameters of the target recommendation models are different. A target recommendation model corresponding to the target scene needs to be determined according to the target scene corresponding to the target user.
It should be noted that the target recommendation model is generated by using one or more algorithms of a collaborative filtering algorithm, a matrix decomposition algorithm, and a neural network matrix decomposition algorithm.
Wherein the recommendation based on the collaborative filtering algorithm is a recommendation based on a user behavior. Collaborative filtering that relies on past behavior of the user. The behavior may be past transaction behavior, product score, and the like. The collaborative filtering algorithm does not require explicit attribute information. Collaborative filtering recommends relationships for new users' products by analyzing the inherent relationships of users and goods.
Two main approaches in the field of collaborative filtering algorithms are nearest neighbor methods and potential factor models. The nearest neighbor method mainly focuses on the product relationship or the user relationship, and is a filtering engine for comparison basis. The latent factor model does not select all relationships, but decomposes the co-occurrence matrix by a matrix decomposition technique, such as extracting 20-100 factors to represent the original matrix information.
The matrix decomposition method is to take users as one dimension, construct a two-dimensional matrix and convert the two-dimensional matrix into a high latitude representation, which may have 50 or 100 dimensions, and the number of the two-dimensional matrix is the same as the size of the content, for example, each content has a small representation of 100 dimensions, and each user also has a small representation of 100 dimensions.
The basic principle of neural network matrix decomposition algorithms considers the model as a black box, which is used to train a representation of a user and a representation of content. Different from matrix decomposition, the neural network matrix decomposition algorithm is realized based on a neural network, and more degrees of freedom are brought. Compared with a matrix decomposition method, the expression capability of the neural network matrix decomposition algorithm is enhanced.
It should be noted that the target recommendation model may be selected according to a scene corresponding to the target user, where the target recommendation model is a model trained in advance.
S205: acquiring a target product of a target scene by using a user hierarchy and a target recommendation model of the target user in the target scene, and recommending product information corresponding to the target product to the target user; the target recommendation model is used for outputting target products corresponding to the target scene according to user hierarchy of the target scene.
And recommending related product information according to user hierarchy of the target user in the target scene. For example, if the user hierarchy in the consumption dimension hierarchy of the user in any consumption scenario is a hierarchy with poor consumption capability and low consumption quality, the product information recommended to the target user and output by the target recommendation model is a study-aid loan, a study-aid fund and the like, or a social practice activity, a public welfare activity and the like, which can be used for improving the credit level and reducing part of the repayment amount of the target user. And for target users with stronger consumption capacity and higher consumption quality, the product information recommended to the target users and output by the target recommendation model is a student credit card system.
Based on the above, according to the information recommendation method provided by the embodiment of the application, the user feature data of the target user is obtained. And inputting the user characteristic data into the user model to obtain user images of the target user in different scenes. Based on the user image of the target user in the target scene, the user hierarchy of the target user in the target scene may be determined. Based on the user hierarchy of the target user in the target scene, a target recommendation model corresponding to the target scene can be determined. And outputting the target product in the target scene by using the user hierarchy of the target user in the target scene and the target recommendation model. And recommending the target product information in the target scene to the target user. Therefore, the target user does not need to provide clear requirements every time, product information capable of meeting the requirements of the target user is automatically recommended to the target user, and the use experience of the user is improved.
In addition, before the user model is used to process the user feature data of the target user, training of the user model is required. In a possible implementation manner, an embodiment of the present application provides a specific implementation manner of a training process of a user model, including:
a1: and acquiring historical user characteristic data and a label corresponding to the historical user characteristic data.
The historical user characteristic data comprises historical user characteristic data of different users. The labels are used for representing different user portraits corresponding to different users in different scenes. The larger the data size of the historical user characteristic data is, the more accurate the output result of the user model obtained by final training is.
The labels corresponding to the historical characteristic data need to determine corresponding different labels according to different historical characteristic data. Tags are, for example, personal hobbies, consumption habits, campus borrowing habits, financial service usage habits, and the like.
A2: and training the user model according to the historical user characteristic data and the labels corresponding to the historical user characteristic data until a preset condition is reached, and acquiring the trained user model.
Specifically, historical user characteristic data is input into a user model, a user portrait output by the user model is obtained, the user portrait output by the user is compared with a label corresponding to the historical user characteristic data to generate a training error, model parameters of the user model are adjusted according to the training error, and an updated user model is obtained. And inputting the historical user characteristic data into the updated user model to obtain output again until a preset condition is reached, and obtaining the trained user model.
It should be noted that the preset condition may be set according to an actual requirement, for example, a training error of the user model reaches a threshold or a number threshold reached by the training number.
After the user model is trained, the user model needs to be tested by using the verification data and the labels corresponding to the verification data, and the model is adjusted and optimized, so that the output result of the user model is more accurate and more accords with the actual situation. The historical user feature data in a1 is training data. The verification data is another batch of historical user characteristic data used for user model verification.
In addition, before acquiring the user characteristic data of the target user, acquiring raw data of the target user is also included. The original data of the target user has null value, illegal data and the like. Based on this, in one possible implementation, the application provides another information recommendation method. Before obtaining the user characteristic data of the target user, another information recommendation method further comprises the following steps:
b1: and acquiring original data of a target user.
For example, the original data of the target user is unprocessed data stored in the campus multi-scenario one-card service platform. The raw data contains attributes such as behavior logs of the users, basic attributes of the users and the like.
B2: and carrying out data cleaning processing and data derivation processing on the original data of the target user to generate user characteristic data of the target user.
And performing data preprocessing on the original data of the target user, wherein the data preprocessing mainly comprises data cleaning processing and data derivation processing. For example: the data cleaning process is to filter out irrelevant fields in the table, specify field types, process default values and the like.
In specific implementation, the data cleaning process mainly comprises:
(1) and (3) checking a legal value: and setting different check values aiming at different scenes, and setting a legal range according to the check values. And forcibly setting data which is not in a legal range in the user characteristic data as a maximum value or judging the data to be invalid, and removing the data. For example, the check value of the birth date is set to 2015 year 5 month 5 day, and the legal range is set to be less than 2015 year 5 month 5 day. The birth date of 5 months 5 days or more from 2015 year included in the user characteristic data is regarded as invalid data.
(2) Null value processing and abnormal value processing: and flexible processing modes such as mean value filling, median filling and the like are adopted according to actual conditions.
In specific implementation, the data derivation processing is to generate derived feature data based on the cleaned raw data. The user characteristic data of the target user comprises derived characteristic data.
It should be noted that the historical user feature data used for training and verifying the user model is also obtained by performing data cleaning processing and data derivation processing on historical raw data.
In practical application, the obtained original data of the target user or the historical original data for training and verifying the user model usually only contain a small amount of data corresponding to basic dimensionality, and are not suitable for directly inputting the model, such as a user address, a daily consumption amount of the user and the like. After data corresponding to the basic dimensions are properly transformed or combined, derived feature data can be generated, the data expansion effect is achieved, and great help is achieved for user model training, verification and application. For example, the user's preference for the product may be inferred from the user's purchase record, and the user's preference for the product may be used as the derived feature data.
It can be understood that the feature dimension of the collected data is not very large, and only data corresponding to a small amount of basic dimensions are included. And the directly collected features cannot fully reflect all information of the user. Therefore, the original data needs to be transformed, combined and extrapolated to generate new derived feature data to augment the data. In addition, the influence magnitude of the derived feature data on the original data can be obtained by a feature importance evaluation method.
In addition, in order to enable the user model and the target recommendation model to be more accurate, after the multi-scene one-card service platform obtains user characteristic data of a new user, the user model and the target recommendation model are trained again by using the newly obtained user characteristic data and historical user characteristic data, and iterative optimization of the model is carried out to obtain the more accurate model.
In addition, in the embodiment of the application, the user characteristic data, the user model and the target recommendation model are displayed on a background interface, so that the operator can conveniently perform related operation.
Based on the information recommendation method provided by the method embodiment, an embodiment of the present application further provides an information recommendation apparatus, which will be described below with reference to the accompanying drawings.
Referring to fig. 3, fig. 3 is a schematic structural diagram of an information recommendation device according to an embodiment of the present application. As shown in fig. 3, the information recommendation apparatus includes:
a first obtaining unit 301, configured to obtain user characteristic data of a target user;
an input unit 302, configured to input the user feature data into a user model, and obtain user images of the target user in different scenes; the user model is used for acquiring user images of the target user in different scenes according to the user characteristic data of the target user;
a second obtaining unit 303, configured to obtain, according to the user image of the target user in the target scene, a user hierarchy of the target user in the target scene; the target scene is any one of the different scenes;
a determining unit 304, configured to determine, according to a target scene corresponding to the target user, a target recommendation model corresponding to the target scene;
a recommending unit 305, configured to obtain a target product of a target scene by using the user hierarchy of the target user in the target scene and the target recommendation model, and recommend product information corresponding to the target product to the target user; the target recommendation model is used for outputting target products corresponding to the target scene according to user hierarchy of the target scene.
In one possible implementation, the apparatus further includes: a model training unit;
the model training unit comprises:
the acquisition subunit is used for acquiring historical user characteristic data and a label corresponding to the historical user characteristic data;
and the training subunit is used for training the user model according to the historical user characteristic data and the labels corresponding to the historical user characteristic data until a preset condition is reached, and acquiring the trained user model.
In one possible implementation, the apparatus further includes:
a third obtaining unit, configured to obtain original data of a target user before obtaining user feature data of the target user;
and the processing unit is used for carrying out data cleaning processing and data derivation processing on the original data of the target user to generate user characteristic data of the target user.
In one possible implementation, the user hierarchy includes one or more of a population dimension hierarchy, a credit dimension hierarchy, a consumption dimension hierarchy, and an interest dimension hierarchy.
The embodiment of the application provides an information recommendation device, which is used for acquiring user characteristic data of a target user. And inputting the user characteristic data into the user model to obtain user images of the target user in different scenes. Based on the user image of the target user in the target scene, the user hierarchy of the target user in the target scene may be determined. Based on the user hierarchy of the target user in the target scene, a target recommendation model corresponding to the target scene can be determined. And outputting the target product in the target scene by using the user hierarchy of the target user in the target scene and the target recommendation model. And recommending the target product information in the target scene to the target user. The device of the embodiment of the application achieves the purposes of acquiring the target product information under different scenes according to the user characteristics of the target user and automatically recommending the target product information to the target user.
In addition, the embodiment of the application also provides an information recommendation system, the information recommendation system at least comprises a campus multi-scene one-card service platform, and the campus multi-scene one-card service platform executes the information recommendation method.
The campus multi-scene one-card service platform provides public resources of colleges and universities, and target users can use the public resources of colleges and universities, such as libraries, databases and laboratories, through the campus multi-scene one-card service platform, so that the resource utilization rate of the public resources of colleges and universities is improved.
As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that all or part of the steps in the above embodiment methods can be implemented by software plus a necessary general hardware platform. Based on such understanding, the technical solution of the present application may be essentially or partially implemented in the form of a software product, which may be stored in a storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the method according to the embodiments or some parts of the embodiments of the present application.
It should be noted that, in the present specification, the embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments may be referred to each other. The method disclosed by the embodiment corresponds to the system disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the system part for description.
It should also be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. An information recommendation method, characterized in that the method comprises:
acquiring user characteristic data of a target user;
inputting the user characteristic data into a user model, and acquiring user images of the target user in different scenes; the user model is used for acquiring user images of the target user in different scenes according to the user characteristic data of the target user;
acquiring user hierarchy of the target user in the target scene according to the user image of the target user in the target scene; the target scene is any one of the different scenes;
determining a target recommendation model corresponding to the target scene according to the target scene corresponding to the target user;
acquiring a target product of a target scene by using the user hierarchy of the target user in the target scene and the target recommendation model, and recommending product information corresponding to the target product to the target user; the target recommendation model is used for outputting target products corresponding to the target scene according to user hierarchy of the target scene.
2. The method of claim 1, wherein the training process of the user model comprises:
acquiring historical user characteristic data and a label corresponding to the historical user characteristic data;
and training a user model according to the historical user characteristic data and the label corresponding to the historical user characteristic data until a preset condition is reached, and acquiring the trained user model.
3. The method of claim 1, wherein prior to said obtaining user characteristic data of the target user, the method further comprises:
acquiring original data of a target user;
and carrying out data cleaning processing and data derivation processing on the original data of the target user to generate user characteristic data of the target user.
4. The method of claim 1, wherein model parameters of the goal recommendation model are different when the goal scenarios are different.
5. The method of claim 1, wherein the user hierarchy includes one or more of a demographic dimension hierarchy, a credit dimension hierarchy, a consumption dimension hierarchy, and an interest and hobby dimension hierarchy.
6. An information recommendation apparatus, characterized in that the apparatus comprises:
the first acquisition unit is used for acquiring user characteristic data of a target user;
the input unit is used for inputting the user characteristic data into a user model and acquiring user images of the target user in different scenes; the user model is used for acquiring user images of the target user in different scenes according to the user characteristic data of the target user;
the second acquisition unit is used for acquiring the user hierarchy of the target user in the target scene according to the user image of the target user in the target scene; the target scene is any one of the different scenes;
the determining unit is used for determining a target recommendation model corresponding to the target scene according to the target scene corresponding to the target user;
the recommendation unit is used for acquiring a target product of a target scene by using the user hierarchy of the target user in the target scene and the target recommendation model, and recommending product information corresponding to the target product to the target user; the target recommendation model is used for outputting target products corresponding to the target scene according to user hierarchy of the target scene.
7. The apparatus of claim 6, further comprising: a model training unit;
the model training unit comprises:
the acquisition subunit is used for acquiring historical user characteristic data and a label corresponding to the historical user characteristic data;
and the training subunit is used for training the user model according to the historical user characteristic data and the labels corresponding to the historical user characteristic data until a preset condition is reached, and acquiring the trained user model.
8. The apparatus of claim 6, further comprising:
a third obtaining unit, configured to obtain original data of a target user before obtaining user feature data of the target user;
and the processing unit is used for carrying out data cleaning processing and data derivation processing on the original data of the target user to generate user characteristic data of the target user.
9. The apparatus of claim 6, wherein the user hierarchy comprises one or more of a demographic dimension hierarchy, a credit dimension hierarchy, a consumption dimension hierarchy, and an interest and hobby dimension hierarchy.
10. An information recommendation system, characterized in that the information recommendation system at least comprises a campus multi-scenario one-card service platform, and the campus multi-scenario one-card service platform executes the information recommendation method according to any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110850844.5A CN113487389A (en) | 2021-07-27 | 2021-07-27 | Information recommendation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110850844.5A CN113487389A (en) | 2021-07-27 | 2021-07-27 | Information recommendation method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113487389A true CN113487389A (en) | 2021-10-08 |
Family
ID=77944061
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110850844.5A Pending CN113487389A (en) | 2021-07-27 | 2021-07-27 | Information recommendation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113487389A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114661994A (en) * | 2022-03-28 | 2022-06-24 | 徐勇 | User interest data processing method and system based on artificial intelligence and cloud platform |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111143697A (en) * | 2020-01-02 | 2020-05-12 | 腾讯科技（深圳）有限公司 | Content recommendation method and related device |
| CN111753026A (en) * | 2020-06-28 | 2020-10-09 | 中国银行股份有限公司 | User portrait generation system, method, device, equipment and medium |
| CN111966835A (en) * | 2020-08-26 | 2020-11-20 | 中国银行股份有限公司 | Device and method for analyzing functional service required by scene based on knowledge graph |
| CN112116448A (en) * | 2020-09-09 | 2020-12-22 | 中国农业银行股份有限公司湖南省分行 | Accurate marketing construction method for three-agriculture client portrait |
-
2021
- 2021-07-27 CN CN202110850844.5A patent/CN113487389A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111143697A (en) * | 2020-01-02 | 2020-05-12 | 腾讯科技（深圳）有限公司 | Content recommendation method and related device |
| CN111753026A (en) * | 2020-06-28 | 2020-10-09 | 中国银行股份有限公司 | User portrait generation system, method, device, equipment and medium |
| CN111966835A (en) * | 2020-08-26 | 2020-11-20 | 中国银行股份有限公司 | Device and method for analyzing functional service required by scene based on knowledge graph |
| CN112116448A (en) * | 2020-09-09 | 2020-12-22 | 中国农业银行股份有限公司湖南省分行 | Accurate marketing construction method for three-agriculture client portrait |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114661994A (en) * | 2022-03-28 | 2022-06-24 | 徐勇 | User interest data processing method and system based on artificial intelligence and cloud platform |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104866969A (en) | Personal credit data processing method and device | |
| Huang et al. | LSCD: Low-rank and sparse cross-domain recommendation | |
| US9665695B2 (en) | Systems and methods for ranking rules that identify potentially illegitimate activities | |
| Helles et al. | Infrastructures of tracking: Mapping the ecology of third-party services across top sites in the EU | |
| CN111611496A (en) | Product recommendation method and device | |
| KR20080019392A (en) | Document auto editing system and the method thereof | |
| CN110717758B (en) | Abnormal transaction identification method and device | |
| CN112016850A (en) | Service evaluation method and device | |
| CN113487389A (en) | Information recommendation method and device | |
| WO2019095846A1 (en) | Shared product recommendation method and apparatus, and electronic device | |
| US9525687B2 (en) | Template for customer attributes | |
| KR102557299B1 (en) | A method of creating the right marketing strategy for your users | |
| CN116756281A (en) | Knowledge question-answering method, device, equipment and medium | |
| CN116467525A (en) | Recommendation method, device, equipment and storage medium of business products | |
| CN110163761B (en) | Suspicious item member identification method and device based on image processing | |
| JP4962950B2 (en) | Recommendation method, recommendation server and program for users on network | |
| Niu et al. | A Coupled User Clustering Algorithm Based on Mixed Data for Web‐Based Learning Systems | |
| OECD | The relationship between local content, Internet development and access prices | |
| DeMars | Big data | |
| Huang et al. | Maverick matters: Client contribution and selection in federated learning | |
| CN112818235A (en) | Violation user identification method and device based on associated features and computer equipment | |
| Dienes | Info Capacity| A Meta Study of 26 “How Much Information” Studies: Sine Qua Nons and Solutions | |
| KR102480239B1 (en) | A method of managing and matching merchant for user | |
| Riederer | Location Data: Perils, Profits, Promise | |
| KR102538748B1 (en) | System for providing photo reward service |
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 |