CN117593089A - Credit card recommendation method, apparatus, device, storage medium and program product - Google Patents

Credit card recommendation method, apparatus, device, storage medium and program product Download PDF

Info

Publication number
CN117593089A
CN117593089A CN202311651204.7A CN202311651204A CN117593089A CN 117593089 A CN117593089 A CN 117593089A CN 202311651204 A CN202311651204 A CN 202311651204A CN 117593089 A CN117593089 A CN 117593089A
Authority
CN
China
Prior art keywords
credit card
recommendation
user
information
historical
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
Application number
CN202311651204.7A
Other languages
Chinese (zh)
Inventor
李云龙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Agricultural Bank of China
Original Assignee
Agricultural Bank of China
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Agricultural Bank of China filed Critical Agricultural Bank of China
Priority to CN202311651204.7A priority Critical patent/CN117593089A/en
Publication of CN117593089A publication Critical patent/CN117593089A/en
Pending legal-status Critical Current

Links

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
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2457—Query processing with adaptation to user needs
    • G06F16/24578—Query processing with adaptation to user needs using ranking
    • 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
    • G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Technology Law (AREA)
  • Computational Linguistics (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The embodiment of the application provides a credit card recommendation method, a device, equipment, a storage medium and a program product, and relates to the technical field of credit card recommendation, wherein the method comprises the following steps: acquiring demand information of a target user, historical data of a historical user and information of each credit card product; determining a first recommendation list of credit card products, a second recommendation list of credit card products and a third recommendation list of credit card products through a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an article-based collaborative filtering recommendation algorithm respectively; and determining a target credit card product recommended for the target user through the first recommendation list, the second recommendation list and the third recommendation list based on the requirement information of the target user. The method provided by the embodiment of the application can solve the problem that the prior art cannot be suitable for diversified scenes to accurately recommend credit card products to users.

Description

Credit card recommendation method, apparatus, device, storage medium and program product
Technical Field
The embodiment of the application relates to the technical field of credit card recommendation, in particular to a credit card recommendation method, a credit card recommendation device, credit card recommendation equipment, a credit card recommendation storage medium and a credit card recommendation program product.
Background
With the development of the internet, the once-lacking information data has been explosively increased as an exponential function. For a user, excessive redundant data can interfere with the retrieval of related information, so that the efficiency of acquiring useful data is greatly reduced; each large information platform can not be positioned to a precise target user due to data flood. This phenomenon also occurs in the credit card product system of banks, and how to enable business personnel to quickly offer customers the credit card products that are suitable and they may like becomes key to research.
At present, a recommendation system is developed to filter a large amount of redundant information and data for users, and interference of the overload information is reduced, so that people can search information more conveniently and rapidly, and user experience effect is improved.
However, the existing recommendation system cannot be suitable for diversified scenes, and credit card products cannot be accurately recommended to users, so that user experience is affected.
Disclosure of Invention
The embodiment of the application provides a credit card recommendation method, device, equipment, storage medium and program product, which are used for solving the problem that the prior art cannot be suitable for diversified scenes to accurately recommend credit card products to users.
In a first aspect, an embodiment of the present application provides a credit card recommendation method, including:
acquiring demand information of a target user, historical data of a historical user and information of each credit card product;
according to the demand information of the target user, the historical data of the historical user and the information of each credit card product, determining a first recommendation list of the credit card product corresponding to the content-based recommendation algorithm, a second recommendation list of the credit card product corresponding to the user-based collaborative filtering recommendation algorithm and a third recommendation list of the credit card product corresponding to the article-based collaborative filtering recommendation algorithm through a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an article-based collaborative filtering recommendation algorithm;
and determining a target credit card product recommended for the target user through the first recommendation list, the second recommendation list and the third recommendation list based on the requirement information of the target user.
In one possible design, the historical data includes historical user demand information, information of credit card products used, and evaluation information of credit card products; according to the demand information of the target user, the history data of the history user and the information of each credit card product, determining a first recommendation list of the credit card product corresponding to the recommendation algorithm based on the content, a second recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering, and a third recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering through a recommendation algorithm based on the content, a collaborative filtering recommendation algorithm based on the user and a collaborative filtering recommendation algorithm based on the object respectively, wherein the method comprises the following steps:
Obtaining content feature vectors of the credit card products through a TF-IDF model according to the information of the credit card products, determining recommendation lists corresponding to the credit card products through a content-based recommendation algorithm according to the content feature vectors of the credit card products, and determining a first recommendation list matched with the requirement information of the target user from the recommendation lists corresponding to the credit card products according to the requirement information of the target user and the historical data of the historical user;
determining a recommendation list corresponding to the credit card products recommended by the historical user through a collaborative filtering recommendation algorithm based on the user according to the historical data of the historical user and the information of each credit card product, and determining a second recommendation list matched with the demand information of the target user from the recommendation list corresponding to the credit card products recommended by the historical user according to the demand information of the target user and the historical data of the historical user;
according to the historical data of the historical users and the information of the credit card products, determining the similarity between the credit card products based on a collaborative filtering recommendation algorithm of the articles, constructing a recommendation list of the similar credit card products, and determining a third recommendation list matched with the requirement information of the target user from the recommendation list of the similar credit card products according to the requirement information of the target user and the historical data of the historical users.
In one possible design, the determining, according to the content feature vector of each credit card product, a recommendation list corresponding to each credit card product through a recommendation algorithm based on content includes:
according to the content feature vector of each credit card product, obtaining the score corresponding to each credit card product through a UGC recommendation model corresponding to a content-based recommendation algorithm; the UGC recommendation model is a recommendation model for adding punishment items to hot tags;
calculating the similarity between different credit card products according to the content feature vector of each credit card product;
and determining a recommendation list corresponding to each credit card product according to the scores corresponding to each credit card product and the similarity between different credit card products, wherein the recommendation list corresponding to each credit card product comprises information of each credit card product which is sequenced based on the scores.
In one possible design, the determining, according to the historical data of the historical user and the information of each credit card product, the recommendation list corresponding to the credit card product recommended to the historical user through a collaborative filtering recommendation algorithm based on the user includes:
According to the historical data of the historical users and the information of each credit card product, determining the similarity between the historical users through a collaborative filtering recommendation algorithm based on the users;
and determining a recommendation list corresponding to the credit card products recommended by the historical users according to the similarity among the historical users, the information of the credit card products used in the historical data of the historical users and the evaluation information of the credit card products.
In one possible design, the building of a recommendation list for similar credit card products includes:
according to the historical data of the historical users and the information of each credit card product, scoring data in a preset time period is obtained from a Redis cache database, wherein the scoring data are used for representing scoring data of the users on the credit card products;
constructing a recommendation list for similar credit card products according to the scoring data of each credit card product; and the recommendation list of each credit card product comprises information of other credit card products similar to the credit card product after being ranked based on the scoring data.
In one possible design, the determining, based on the requirement information of the target user, the target credit card product recommended to the target user through the first recommendation list, the second recommendation list and the third recommendation list includes:
According to the demand information of the target user, determining weights respectively corresponding to the first recommendation list, the second recommendation list and the third recommendation list in the current service scene;
determining a target credit card product recommended for the target user according to weights respectively corresponding to the first recommendation list, the second recommendation list and the third recommendation list;
correspondingly, the method further comprises the steps of:
acquiring feedback information of the target user on the recommended target credit card product;
and updating the historical data of the historical user and the information of each credit card product according to the feedback information.
In a second aspect, an embodiment of the present application provides a credit card recommendation device, including:
the acquisition module is used for acquiring the demand information of the target user, the historical data of the historical user and the information of each credit card product;
the recommendation module is used for determining a first recommendation list of credit card products corresponding to the recommendation algorithm based on the content, a second recommendation list of credit card products corresponding to the collaborative filtering recommendation algorithm based on the user and a third recommendation list of credit card products corresponding to the collaborative filtering recommendation algorithm based on the article according to the demand information of the target user, the historical data of the historical user and the information of each credit card product through the recommendation algorithm based on the content, the collaborative filtering recommendation algorithm based on the user and the collaborative filtering recommendation algorithm based on the article;
And the recommending module is also used for determining a target credit card product recommended for the target user through the first recommending list, the second recommending list and the third recommending list based on the requirement information of the target user.
In a third aspect, an embodiment of the present application provides an electronic device, including: at least one processor and memory;
the memory stores computer-executable instructions;
the at least one processor executes computer-executable instructions stored in the memory, causing the at least one processor to perform the credit card recommendation method as described above in the first aspect and the various possible designs of the first aspect.
In a fourth aspect, embodiments of the present application provide a computer-readable storage medium, where computer-executable instructions are stored, and when executed by a processor, implement the credit card recommendation method according to the first aspect and the various possible designs of the first aspect.
In a fifth aspect, embodiments of the present application provide a computer program product comprising a computer program which, when executed by a processor, implements the credit card recommendation method according to the first aspect and the various possible designs of the first aspect.
The credit card recommendation method, device, equipment, storage medium and program product provided in this embodiment firstly acquire demand information of a target user, history data of a history user and information of each credit card product; then according to the demand information of the target user, the historical data of the historical user and the information of each credit card product, determining a first recommendation list of the credit card product corresponding to the recommendation algorithm based on the content, a second recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering of the user and a third recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering of the article through a recommendation algorithm based on the content, a collaborative filtering recommendation algorithm based on the user and a collaborative filtering recommendation algorithm based on the article; and determining a target credit card product recommended for the target user through the first recommendation list, the second recommendation list and the third recommendation list based on the requirement information of the target user. Therefore, by acquiring the demand information of the user and combining different recommendation algorithms, the demands of the user are comprehensively analyzed, one or more credit card products suitable for the user or liked by the user are recommended for the user, the method is suitable for diversified scenes under different demands of different users, matched credit card products can be accurately recommended for the user, the personalized demands of the user are met, and user experience is further improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief description will be given below of the drawings that are needed in the embodiments or the prior art descriptions, it being obvious that the drawings in the following description are some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort to a person skilled in the art.
Fig. 1 is a schematic view of a scenario of a credit card recommendation method provided in an embodiment of the present application;
FIG. 2 is a flowchart illustrating a credit card recommendation method according to an embodiment of the present disclosure;
FIG. 3 is a flowchart illustrating a credit card recommendation method according to an embodiment of the present disclosure;
FIG. 4 is a schematic structural diagram of a credit card recommendation device according to an embodiment of the present disclosure;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
For the purposes of making the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application.
The terms "first," "second," "third," "fourth" and the like in the description and in the claims of this application and in the above-described figures, if any, 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 embodiments of the present application described herein may be implemented, for example, 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, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be noted that, the user information (including but not limited to user equipment information, user personal information, etc.) and the 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 fully authorized by each party, and the collection, use and processing of the related data need to comply with the related laws and regulations and standards, and provide corresponding operation entries for the user to select authorization or rejection.
At present, a recommendation system is developed to filter a large amount of redundant information and data for users, and interference of the overload information is reduced, so that people can search information more conveniently and rapidly, and user experience effect is improved. However, the existing recommendation system cannot be suitable for diversified scenes, and credit card products cannot be accurately recommended to users, so that user experience is affected.
Aiming at the problems existing in the prior art, the technical concept of the method is to comprehensively analyze the demands of the users by acquiring the demand information of the users and combining different recommendation algorithms, recommend one or more credit card products suitable for the users or liked by the users for the users, be applicable to diversified scenes under different demands of different users, accurately recommend matched credit card products for the users, meet the personalized demands of the users and further improve the user experience.
Term interpretation:
content-based recommendation algorithm: items are defined by their own associated feature attributes and the degree of similarity between the various items before can be found, and then the items with the highest degree of similarity are recommended to the user based on the user's historical data (e.g., which credit card products have been used or liked (e.g., labeled), etc.).
Collaborative filtering recommendation algorithm based on user: the user's preference for items is obtained by mining the user's historical data, then searching for a group of nearby users that have similar interests as the current user, and recommending their preferred items to the current user.
Collaborative filtering recommendation algorithm based on articles: for example, product A is liked by user A and user B; product B is also liked by user a and user B, then product a and product B are considered similar, and when user C likes product a, then user C is considered to also like product B, then product B is also recommended to user C.
A lingering semantic model: the latent meaning analysis technology can be used for mining potential relations between different sentences and texts, namely filling the original scoring matrix through the dimension reduction technology.
Alternate least squares: after the matrix P and the matrix Q are coupled together by matrix multiplication, in order to decouple them, the matrix Q may be obtained by fixing the matrix Q in advance, taking the other matrix P as a variable, and minimizing the loss function, and likewise, the matrix P may be obtained by fixing the matrix P in advance, taking the other matrix Q as a variable, and minimizing the loss function.
Word frequency-inverse document frequency model (TF-IDF): the method is a weighting technology for text mining and information retrieval, and can evaluate the importance degree of a word in a specified document in a document set or corpus. The importance of the word increases with increasing number of occurrences in the document and also decreases with increasing number of occurrences in the corpus. The formula is as follows:
TF-DF＝TF×IDF
flume-ng: a system for collecting log data can efficiently collect and aggregate a large amount of log data obtained from different data sources and store the obtained data in a specific data storage system.
Redis cache database: redis is a key-value database that is extremely high in performance and capable of supporting rich data types, and because it is itself memory-based, the overhead required to read data from disk to content is eliminated compared to disk data.
Kafka: a distributed publishing and subscribing message system with extremely high throughput can complete the construction of real-time stream data pipelines among the systems, and can not only read, write and store data streams, but also realize real-time processing of stream data.
In practical application, referring to fig. 1, fig. 1 is a schematic view of a scenario of a credit card recommendation method according to an embodiment of the present application. The execution subject of the present application may be a credit card recommender, which may be deployed in an electronic device, such as a terminal device or a server.
Illustratively, taking a credit card recommendation scenario as an example, the scenario may include: a user terminal 101 and a banking system terminal 102; the user terminal is used to provide a search function for recommending credit card products, and the banking system terminal 102 is equipped with a credit card recommendation device. When a user has a demand, the demand information (such as information of one or more dimensions of the use, appearance, amount, etc. of the credit card needed) can be input into a search box corresponding to the search function through a search function provided by the user terminal 101, the user terminal 101 reports the demand information to the banking system terminal 102, and the credit card recommendation device of the banking system terminal 102 obtains the historical data of the historical user and the stored information (such as product name, card brand, card validity period, amount, provided service, etc. in the credit card product) of each credit card product from the database according to the obtained demand information of the user, and inputs the information into the recommendation model to obtain the target credit card product recommended for the user.
The recommendation model is a model based on fusion of multiple recommendation algorithms, wherein the multiple recommendation algorithms comprise a content-based recommendation algorithm and a collaborative filtering recommendation algorithm based on a latent meaning model, and the collaborative filtering recommendation algorithm based on the latent meaning model can comprise a collaborative filtering recommendation algorithm based on a user, a collaborative filtering recommendation algorithm based on an article and the like. Then, the recommended credit card products or the recommended list of the credit card products recommended by each recommendation algorithm are used for determining the recommended target credit card products for the user through weighted calculation.
The problem that a single recommendation algorithm can not accurately recommend proper credit card products for users aiming at diversified scenes is solved, namely:
1. the recommendation algorithm based on the content requires the complete and comprehensive model as much as possible in the modeling process of the content, and can perfectly express the content of the recommendation algorithm based on the content, but classification labels are mostly used for describing the content in a push system based on the content, so that the one-sided property of the description is caused. The process of extracting the attribute needs to contain interpretable meaning, so that the extraction maintenance needs to be performed manually, thereby causing the waste of human resources. The content-based recommendation algorithm adopted in the embodiment of the application fuses the TF-IDF model to obtain the content feature vector of each credit card product, and the problem of popular label list frequently encountered in the UGC-based recommendation model can be corrected through the model, so that the recommendation accuracy is improved.
2. The collaborative filtering recommendation algorithm based on the users has the advantages that when the number of the users is far greater than the number of the articles, the cost for calculating the similarity matrix of the users is overlarge, and the calculation amount is greatly increased along with the dynamic increase of the number of the users. Thus, there is another recommended way to: when the number of users reaches a certain threshold, the weight of the users can be set to 0 or the recommendation algorithm is not adopted, and other recommendation algorithms, such as collaborative filtering recommendation algorithm based on articles, are adopted again to be used in combination with the recommendation algorithm based on contents.
3. The collaborative filtering recommendation algorithm based on the article is too dependent on historical data, so that the problems of cold start and data sparsity exist, and therefore, the collaborative filtering recommendation algorithm based on the article can be combined with other algorithms (such as a recommendation algorithm based on content and/or a collaborative filtering recommendation algorithm based on a user, and the like) to solve the problems of cold start and data sparsity.
Therefore, by acquiring the demand information of the user and combining different recommendation algorithms, the demands of the user are comprehensively analyzed, one or more credit card products suitable for the user or liked by the user are recommended for the user, the method is suitable for diversified scenes under different demands of different users, matched credit card products can be accurately recommended for the user, the personalized demands of the user are met, and user experience is further improved.
The technical scheme of the present application is described in detail below with specific examples. The following embodiments may be combined with each other, and some embodiments may not be repeated for the same or similar concepts or processes.
Fig. 2 is a flow chart of a credit card recommendation method according to an embodiment of the present application, where the method may include:
s201, obtaining demand information of a target user, historical data of a historical user and information of each credit card product.
In this embodiment, the execution subject of the credit card recommendation method may be a credit card recommendation apparatus, which may be deployed in an electronic device, such as a terminal device, a server, or the like. Specifically, as shown in fig. 1, when a user has a demand, the demand information (such as information of one or more dimensions of usage, appearance, credit line, etc. of a credit card is required) can be input into a search box corresponding to the search function through a search function provided by the user terminal 101, the user terminal 101 reports the demand information to the banking system terminal 102, and the credit card recommendation device of the banking system terminal 102 obtains the history data of the history user and the stored information (such as product name, card brand, card validity period, credit line, provided service, etc. in the credit card product) of each credit card product from the database according to the obtained demand information of the user.
S202, according to the demand information of the target user, the historical data of the historical user and the information of each credit card product, determining a first recommendation list of the credit card product corresponding to the recommendation algorithm based on the content, a second recommendation list of the credit card product corresponding to the collaborative filtering recommendation algorithm based on the user and a third recommendation list of the credit card product corresponding to the collaborative filtering recommendation algorithm based on the object through a recommendation algorithm based on the content, a collaborative filtering recommendation algorithm based on the user and a collaborative filtering recommendation algorithm based on the object.
In this embodiment, in order to ensure accuracy of recommending credit card products, at least three recommendation algorithms including a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm, and an article-based collaborative filtering recommendation algorithm may be combined to determine a third recommendation list of credit card products corresponding to each recommendation algorithm, and then determine a target credit card product that is finally recommended to the user through weighted calculation.
Yet another recommended way is: determining a recommendation algorithm according to the demand information of the target user, the historical data of the historical user and the information of each credit card product, wherein the recommendation algorithm comprises at least two of a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an article-based collaborative filtering recommendation algorithm, and determining a recommendation list of the credit card products corresponding to each recommendation algorithm according to the demand information of the target user, the historical data of the historical user and the information of each credit card product and through each recommendation algorithm in the recommendation algorithm.
S203, determining a target credit card product recommended for the target user through the first recommendation list, the second recommendation list and the third recommendation list based on the requirement information of the target user.
In this embodiment, according to the requirement information of the target user, the target credit card product recommended for the target user is determined through weighted calculation of the first recommendation list, the second recommendation list and the third recommendation list.
In one possible design, the determining, based on the requirement information of the target user, the target credit card product recommended to the target user through the first recommendation list, the second recommendation list and the third recommendation list may be implemented by:
step a1, determining weights corresponding to the first recommendation list, the second recommendation list and the third recommendation list respectively under the current service scene according to the requirement information of the target user;
and a2, determining a target credit card product recommended by the target user according to weights respectively corresponding to the first recommendation list, the second recommendation list and the third recommendation list.
In this embodiment, according to the requirement information of the target user, the current service scenario and the applicable recommendation algorithm (including at least two of a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an item-based collaborative filtering recommendation algorithm) are determined by combining the historical data of the user and the information of each credit card product, for example: when the data volume of the historical user does not reach a certain threshold, a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an article-based collaborative filtering recommendation algorithm can be adopted for comprehensive analysis, so that matched credit card products are recommended to the user, the recommendation accuracy is ensured, the user requirements are met, and the user experience is improved. When the data volume of the historical user reaches a certain threshold, a combination of a content-based recommendation algorithm and an item-based collaborative filtering recommendation algorithm can be adopted to save computational resources. The selection of the three algorithms is not particularly limited herein, and may depend on the specific scenario or service requirements.
The credit card recommendation method comprises the steps of firstly obtaining demand information of a target user, historical data of a historical user and information of each credit card product; then according to the demand information of the target user, the historical data of the historical user and the information of each credit card product, determining a first recommendation list of the credit card product corresponding to the recommendation algorithm based on the content, a second recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering of the user and a third recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering of the article through a recommendation algorithm based on the content, a collaborative filtering recommendation algorithm based on the user and a collaborative filtering recommendation algorithm based on the article; and determining a target credit card product recommended for the target user through the first recommendation list, the second recommendation list and the third recommendation list based on the requirement information of the target user. Therefore, by acquiring the demand information of the user and combining different recommendation algorithms, the demands of the user are comprehensively analyzed, one or more credit card products suitable for the user or liked by the user are recommended for the user, the method is suitable for diversified scenes under different demands of different users, matched credit card products can be accurately recommended for the user, the personalized demands of the user are met, and user experience is further improved.
In one possible design, the historical data includes historical user demand information and information of credit card products used; according to the demand information of the target user, the historical data of the historical user and the information of each credit card product, the first recommendation list of the credit card product corresponding to the recommendation algorithm based on the content, the second recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering of the user and the third recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering of the article are respectively determined through a recommendation algorithm based on the content, a collaborative filtering recommendation algorithm based on the user and a collaborative filtering recommendation algorithm based on the article, and the method can be realized by the following steps:
step b1, obtaining content feature vectors of the credit card products through a TF-IDF model according to the information of the credit card products, determining recommendation lists corresponding to the credit card products through a recommendation algorithm based on content according to the content feature vectors of the credit card products, and determining a first recommendation list matched with the demand information of the target user from the recommendation lists corresponding to the credit card products according to the demand information of the target user and the historical data of the historical user;
Step b2, determining a recommendation list corresponding to the credit card products recommended for the historical user through a collaborative filtering recommendation algorithm based on the user according to the historical data of the historical user and the information of each credit card product, and determining a second recommendation list matched with the demand information of the target user from the recommendation list corresponding to the credit card products recommended for the historical user according to the demand information of the target user and the historical data of the historical user;
and b3, determining the similarity between the credit card products based on a collaborative filtering recommendation algorithm of the articles according to the historical data of the historical users and the information of the credit card products, constructing a recommendation list of the similar credit card products, and determining a third recommendation list matched with the demand information of the target user from the recommendation list of the similar credit card products according to the demand information of the target user and the historical data of the historical users.
In this embodiment, as shown in fig. 3, the demand information of the target user, the history data of the history user and the information of each credit card product are respectively input into a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an article-based collaborative filtering recommendation algorithm to obtain a recommendation list of the credit card products corresponding to each recommendation algorithm (i.e., a first recommendation list of the credit card products corresponding to the content-based recommendation algorithm, a second recommendation list of the credit card products corresponding to the user-based collaborative filtering recommendation algorithm and a third recommendation list of the credit card products corresponding to the article-based collaborative filtering recommendation algorithm), and then the target credit card products finally recommended to the user are obtained through weighted calculation. Wherein the target credit card product is one or more.
In one possible design, the determining, according to the content feature vector of each credit card product, a recommendation list corresponding to each credit card product through a recommendation algorithm based on content includes:
according to the content feature vector of each credit card product, obtaining the score corresponding to each credit card product through a UGC recommendation model corresponding to a content-based recommendation algorithm; the UGC recommendation model is a recommendation model for adding punishment items to hot tags;
calculating the similarity between different credit card products according to the content feature vector of each credit card product;
and determining a recommendation list corresponding to each credit card product according to the scores corresponding to each credit card product and the similarity between different credit card products, wherein the recommendation list corresponding to each credit card product comprises information of each credit card product which is sequenced based on the scores.
In this embodiment, a mongoDB database is selected for use in storing data. Meanwhile, the data are also loaded into the elastic search server, and through the service, the relevance recommendation when the credit card product is typed can be realized.
Specifically, the content-based recommendation algorithm to be implemented in the present application may treat the product name, card brand, card validity period, etc. in the credit card product as the feature content of each card product, and because the difficulty of directly converting the content into scoring is relatively high, the TF-IDF model is used to obtain the content feature vector of each credit card product, and through the model, the popular tag list problem frequently encountered in the content-based recommendation algorithm (for example, based on UGC recommendation) can be corrected.
Firstly, carrying out preliminary statistics on labels frequently used by each user, then, calculating the article with the largest number of times of the labels for each obtained label, finally, carrying out user article matching, and recommending the article with the most commonly used label of the user to the user, wherein in the algorithm, a scoring formula of a certain user U for the article A is as follows:
from this equation, it can be seen that the score calculated by this method tends to weight more popular tags, and if a popular item is more popular, it will appear in the user's recommendation list, thereby reducing the novelty of the recommendation list. The above mentioned TF-IDF model can better solve this problem, because the IDF term in its formula penalizes the popular label to some extent, where the word frequency corresponds to the frequency of occurrence of a label in all labels of an article, and the inverse document frequency corresponds to the frequency of occurrence of the label in other article label sets, and on the basis of the UGC recommendation model, the TF-IDF model is introduced, so that after adding a certain penalty term to the popular label, the scoring formula is as follows:
Wherein n is u,x Indicating the number of times the user U has marked the label x, n A,x Indicating the number of times the card product a is labeled x,indicating how many users have used a particular tag x, +.>Indicating how many different users have labeled the article a (here, credit card product a, which will not be described in detail) in total, the penalty for hot labels and hot articles is accomplished by the addition of these two items, respectively. Using TF model to characterize content to be extractedThe method comprises the steps of carrying out segmentation vectorization processing, carrying out statistics on segmented type columns according to word frequency, obtaining a plurality of sparse vectors, wherein each sparse vector is composed of three parts, namely vector dimensions, positions corresponding to each label in a feature vector and occurrence frequencies of the sparse vectors, training obtained data by using an IDF model, evaluating the strength of a card product represented by each latitude in the feature vector, and displaying the card product by using an additional features column.
Then, the similarity between the products of each card (here, referred to as a credit card, and not described in detail below) needs to be calculated, so that a recommendation list can be obtained for a single card product, and the similarity between different card products can be described by cosine similarity, and the calculation formula is as follows:
Wherein A and B represent two card products, t Ai Each item, t, in the feature vector representing card product a Bi Each item in the feature vector representing card product B.
After all credit card products are subjected to Cartesian product and a scoring matrix is obtained, the similarity of all the credit card products can be calculated in pairs through a cosine similarity formula, and finally the credit card products are filled into a recommendation list of each credit card product according to the descending order of scores (wherein one credit card product corresponds to one recommendation list).
In one possible design, the determining, according to the historical data of the historical user and the information of each credit card product, the recommendation list corresponding to the credit card product recommended to the historical user through a collaborative filtering recommendation algorithm based on the user includes:
according to the historical data of the historical users and the information of each credit card product, determining the similarity between the historical users through a collaborative filtering recommendation algorithm based on the users;
and determining a recommendation list corresponding to the credit card products recommended by the historical users according to the similarity among the historical users and the information of the credit card products used in the historical data of the historical users.
In one possible design, the building of a recommendation list for similar credit card products includes:
according to the historical data of the historical users and the information of each credit card product, scoring data in a preset time period is obtained from a Redis cache database, wherein the scoring data are used for representing scoring data of the users on the credit card products;
constructing a recommendation list for similar credit card products according to the scoring data of each credit card product; and the recommendation list of each credit card product comprises information of other credit card products similar to the credit card product after being ranked based on the scoring data.
Specifically, a collaborative filtering recommendation algorithm based on a latent meaning model (which may include a collaborative filtering recommendation algorithm based on a user and a collaborative filtering recommendation algorithm based on an article; in order to obtain a feature vector matrix corresponding to the user or a product, and then calculate similarity), an ALS algorithm is required to be used to perform matrix decomposition on a scoring matrix constructed by a scoring table (one scoring for each card product, and the scoring of a plurality of card products forms the scoring table), so as to obtain two feature matrices P and Q, where the two feature matrices imply features of the user and the card product. The steps for obtaining these two matrices are as follows:
(1) An initial value Q0 is given to the matrix Q by a pseudo-random number.
(2) The value of matrix Q0 is fixed and the value of matrix P0 is solved.
(3) The value of matrix P0 is fixed, and the value of matrix Q1 is solved.
(4) The value of matrix Q1 is fixed and the value of matrix P1 is solved.
(5) Repeating the steps (2) to (4).
(6) Until the error value meets a threshold condition, or the number of iterations reaches an upper limit.
If the matrix Q is fixed, the value of the matrix P is obtainedThe process in the row solving operation is divided into the following steps, firstly, for the user feature matrix P, the feature vector P of each user U U Are independent, the obtained values are independent of the feature vectors of the rest of the users, so that each feature vector P can be U With separate solutions, the optimization objective can be transformed into the following formula:
wherein R is Ui Representing the score of a user U for a certain card product i in a scoring matrix, P U Feature vector, Q representing user U i Representing the feature vector of the card product i.
Let the latter half of the formula be L U (P U ) The next task is to find the feature vector P of a user U Can make L U (P U ) Take the minimum value through L U (P U ) P pair P U The bias derivative can be obtained by the following formula:
/>
Let the bias guide be 0, the following formula can be further obtained:
due to P U And Q is equal to i The multiplication results in a number, so that the number can be transposed in order and then exchanged, and P can be further added U Is extracted from the transposed matrix of (a), the resulting formula is shown below:
wherein Q is i Representing the feature vector of the card product, will in fact beThe characteristic matrix of the card product is obtained by splitting according to columns and is Q in the above formula i A summation accumulation section for recovering it to Q and at P U The transposed matrix of (2) is multiplied by the inverse of its original matrix to obtain the user feature vector P U Is the final expression of (2):
through the steps, one P can be solved each time U Finally, the P is calculated U And all are combined to obtain the user characteristic matrix P.
The following can be calculated to obtain the card product feature matrix Q: and fixing the obtained matrix P, repeating the steps to obtain the value of the characteristic matrix Q of the card product, and respectively obtaining the user characteristic matrix P and the credit card characteristic matrix Q after the iteration times reach the upper limit. In order to minimize the error obtained by the ALS model, the parameter adjustment is needed, and in this application, the Root Mean Square Error (RMSE) is taken as an error consideration between the prediction score and the actual score, and the formula is as follows:
Wherein, root Mean Square Error (RMSE) formula is a model for evaluating a recommendation system, true i Representing the actual scoring, prediction, of a user for a certain credit card product, here card product i i Indicating that predictive scoring by the recommendation system is performed.
The credit card feature matrix Q is obtained while the list of credit card products recommended for each user is obtained, and based on this matrix Q, the similarity of different card products can be calculated together, and the cosine similarity is used as a standard for comparison between two card products in the same manner as the similarity calculation in the content-based recommendation algorithm, so that the improved article-based collaborative filtering real-time recommendation algorithm to be mentioned later can be padded with required data in advance. The improved collaborative filtering real-time recommendation algorithm based on the articles can be used for the Kafka service, the Redis cache database and the collaborative filtering recommendation technology based on the articles.
Wherein, the similarity matrix of different card products is calculated by a collaborative filtering recommendation algorithm based on a latent semantic model before, and the original collaborative filtering recommendation algorithm based on articles can be improved according to the near several scores of users. The improvement idea is to integrate recent scoring data stored in the Redis cache database into an article-based collaborative filtering recommendation algorithm to obtain a new recommendation list in consideration of recent scoring of certain card products and consumption habits of users. In order to ensure the true reliability of the data tested after the algorithm is finished, the scores of the users with ID 1 in the original rates table are stored in the Redis cache database in advance, and after the real-time score message stream from Kafka is received, the real-time score message stream is firstly divided according to separators and the data types are converted, and then the data types are unfolded around the following formulas:
Wherein q represents a recommendation list which is most similar to a certain card product after the user scores the card product, namely a recommendation result obtained by a collaborative filtering recommendation algorithm directly based on an article, r is the card product which is evaluated by the user for K times recently, sim (q, r) can use cosine similarity to obtain the similarity degree of the card product q and the card product r, rr represents the score of the user on the card product r, sim_sum is the number of card products with similarity meeting a set threshold, and the last two lg (max { strong,1} and lg (max { weak,1} respectively represent a rewarding item and a punishment item) can play a role when the hobbies of the user have obvious tendencies.
Then, the first step is to obtain the latest K scores of the user by connecting with the Redis, then obtain the recommendation list of the currently scored card products by the algorithm, but here, need to additionally read the rates table to find the card products scored by the user, and exclude the same result from the alternative list to prevent the user from being recommended, so as to calculate the first term of the formula, namely the initial score, calculate the similarity by matching each part of the obtained card product recommendation list and each part of the last K scores one by cycling, multiply the evaluation scores of the user on the currently cycled scored card products, and finally add the terms with the similarity meeting the threshold condition to average. After the processing, if the card product to be recommended is similar to the card product with a certain low score of the user, the priority is obviously improved, and if the card product to be recommended is similar to the card product with a certain low score of the user, the acquired score is reduced. The latter two items can act as punishment rewards when the user's hobbies are obviously biased, for example, the score is set to be more than 4 and particularly biased, the score is set to be extremely dislike less than 2, when the score data conforming to the two settings in the near K scores are read, the recommendation priority is obviously influenced, but in order to prevent the influence from being excessive, the log function is selected to slow down the growth trend.
Finally, the recommendation results of the three recommendation algorithms are mixed in a certain proportion to obtain a target credit card product which is finally recommended to the user, so that the accuracy of recommendation can be ensured, the personalized requirements of the user can be met, and further the user experience is improved. In one possible design, the method further comprises:
acquiring feedback information of the target user on the recommended target credit card product; the feedback information at least comprises scoring data of the recommended result;
and optimizing a recommendation model according to the feedback information, wherein the recommendation model is determined by at least two of a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an item-based collaborative filtering recommendation algorithm.
In this embodiment, after recommending the target credit card product for the user, the scoring data of the user on the recommendation result may be obtained, and the corresponding recommendation model may be continuously optimized based on the scoring data, so as to improve the accuracy of recommendation.
According to the method and the device, the credit card product name, the card brand, the card validity period and the like are extracted and processed as the characteristic content of each card product, and each product is characterized in multiple dimensions, so that the credit card product name, the card brand, the card validity period and the like can be better represented by a data model. By fusing various recommendation algorithms, the shortages of some algorithms can be improved and the advantages of other algorithms can be used for stippling and compensating, and the real-time performance of credit card recommendation can be enhanced by collecting recent scores of users.
Therefore, the problem of popular label list is solved by integrating content feature vectors with various dimensions and combining a TF-IDF model, and a card product can be comprehensively depicted from a plurality of dimensions. The collaborative filtering recommendation algorithm based on the articles and the collaborative filtering recommendation algorithm based on the users are combined, so that the diversity and the novelty of recommendation are ensured to a certain extent, and better real-time performance and expandability are achieved. The whole recommendation model is optimized to a certain degree through the scoring data of the user in the last few times, so that the recommended card products are guaranteed to have strong real-time performance.
In order to implement the credit card recommendation method, the embodiment provides a credit card recommendation device. Referring to fig. 4, fig. 4 is a schematic structural diagram of a credit card recommendation device according to an embodiment of the present application; the credit card recommendation device 40 includes:
the acquisition module is used for acquiring the demand information of the target user, the historical data of the historical user and the information of each credit card product;
a recommendation module 401, configured to determine, according to the requirement information of the target user, the historical data of the historical user, and the information of each credit card product, a first recommendation list of the credit card product corresponding to the recommendation algorithm based on the content, a second recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering, and a third recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering, by using the recommendation algorithm based on the content, the collaborative filtering recommendation algorithm based on the user, and the collaborative filtering recommendation algorithm based on the item;
The recommendation module 402 is further configured to determine, based on the requirement information of the target user, a target credit card product recommended for the target user through the first recommendation list, the second recommendation list, and the third recommendation list.
In this embodiment, by setting the obtaining module 401 and the recommending module 402, the method is used for comprehensively analyzing the demands of the user by obtaining the demand information of the user and combining different recommending algorithms, recommending one or more credit card products suitable for the user or liked by the user for the user, and is suitable for diversified scenes under different demands of different users, so that matched credit card products can be accurately recommended for the user, and personalized demands of the user are met, thereby improving user experience.
The device provided in this embodiment may be used to implement the technical solution of the foregoing method embodiment, and its implementation principle and technical effects are similar, and this embodiment will not be described herein again.
In one possible design, the historical data includes historical user demand information, information of credit card products used, and evaluation information of credit card products; the recommendation module comprises:
a first recommendation unit, configured to obtain, according to information of each credit card product, a content feature vector of each credit card product through a TF-IDF model, determine, according to a content-based recommendation algorithm, a recommendation list corresponding to each credit card product, and determine, according to requirement information of the target user and historical data of a historical user, a first recommendation list matching requirement information of the target user from the recommendation lists corresponding to each credit card product;
A second recommendation unit, configured to determine, according to the historical data of the historical user and the information of each credit card product, a recommendation list corresponding to the credit card product recommended for the historical user through a collaborative filtering recommendation algorithm based on the user, and determine, according to the demand information of the target user and the historical data of the historical user, a second recommendation list matching with the demand information of the target user from the recommendation lists corresponding to the credit card product recommended for the historical user;
and the third recommendation unit is used for determining the similarity between the credit card products based on a collaborative filtering recommendation algorithm of the articles according to the historical data of the historical users and the information of the credit card products, constructing recommendation lists of similar credit card products, and determining a third recommendation list matched with the demand information of the target user from the recommendation lists of the similar credit card products according to the demand information of the target user and the historical data of the historical users.
In one possible design, the first recommendation unit is specifically configured to:
according to the content feature vector of each credit card product, obtaining the score corresponding to each credit card product through a UGC recommendation model corresponding to a content-based recommendation algorithm; the UGC recommendation model is a recommendation model for adding punishment items to hot tags;
Calculating the similarity between different credit card products according to the content feature vector of each credit card product;
and determining a recommendation list corresponding to each credit card product according to the scores corresponding to each credit card product and the similarity between different credit card products, wherein the recommendation list corresponding to each credit card product comprises information of each credit card product which is sequenced based on the scores.
In one possible design, the second recommendation unit is specifically configured to:
according to the historical data of the historical users and the information of each credit card product, determining the similarity between the historical users through a collaborative filtering recommendation algorithm based on the users;
and determining a recommendation list corresponding to the credit card products recommended by the historical users according to the similarity among the historical users and the information of the credit card products used in the historical data of the historical users.
In one possible design, the third recommendation unit is specifically configured to:
according to the historical data of the historical users and the information of each credit card product, scoring data in a preset time period is obtained from a Redis cache database, wherein the scoring data are used for representing scoring data of the users on the credit card products;
Constructing a recommendation list for similar credit card products according to the scoring data of each credit card product; and the recommendation list of each credit card product comprises information of other credit card products similar to the credit card product after being ranked based on the scoring data.
In one possible design, the recommendation module is further specifically configured to:
according to the demand information of the target user, determining weights respectively corresponding to the first recommendation list, the second recommendation list and the third recommendation list in the current service scene;
determining a target credit card product recommended for the target user according to weights respectively corresponding to the first recommendation list, the second recommendation list and the third recommendation list;
correspondingly, the device further comprises: updating a module; an updating module for:
acquiring feedback information of the target user on the recommended target credit card product; the feedback information at least comprises scoring data of the recommended result;
and optimizing a recommendation model according to the feedback information, wherein the recommendation model is determined by at least two of a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an item-based collaborative filtering recommendation algorithm.
In order to implement the credit card recommendation method, the embodiment provides electronic equipment. Fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 5, the electronic device 50 of the present embodiment includes: at least one processor 501 and memory 502; wherein, the memory 502 is used for storing computer execution instructions; at least one processor 501 for executing computer-executable instructions stored in memory to perform the steps described in the embodiments above. Reference may be made in particular to the relevant description of the embodiments of the method described above.
The embodiment of the application also provides a computer readable storage medium, wherein computer execution instructions are stored in the computer readable storage medium, and when a processor executes the computer execution instructions, the credit card recommendation method is realized.
Embodiments of the present application also provide a computer program product comprising a computer program which, when executed by a processor, implements a credit card recommendation method as described above.
In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described embodiments of the apparatus are merely illustrative, and for example, the division of the modules is merely a logical function division, and there may be additional divisions when actually implemented, for example, multiple modules may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, 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 modules, which may be in electrical, mechanical, or other forms. In addition, each functional module in each embodiment of the present application may be integrated in one processing unit, or each module may exist alone physically, or two or more modules may be integrated in one unit. The units formed by the modules can be realized in a form of hardware or a form of hardware and software functional units.
The integrated modules, which are implemented in the form of software functional modules, may be stored in a computer readable storage medium. The software functional module is stored in a storage medium, and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (english: processor) to perform some of the steps of the methods described in the embodiments of the present application. It should be understood that the above processor may be a central processing unit (english: central Processing Unit, abbreviated as CPU), or may be other general purpose processors, digital signal processors (english: digital Signal Processor, abbreviated as DSP), application specific integrated circuits (english: application Specific Integrated Circuit, abbreviated as ASIC), or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the present invention may be embodied directly in a hardware processor for execution, or in a combination of hardware and software modules in a processor for execution.
The memory may comprise a high-speed RAM memory, and may further comprise a non-volatile memory NVM, such as at least one magnetic disk memory, and may also be a U-disk, a removable hard disk, a read-only memory, a magnetic disk or optical disk, etc. The bus may be an industry standard architecture (Industry Standard Architecture, ISA) bus, an external device interconnect (Peripheral Component, PCI) bus, or an extended industry standard architecture (Extended Industry Standard Architecture, EISA) bus, among others. The buses may be divided into address buses, data buses, control buses, etc. For ease of illustration, the buses in the drawings of the present application are not limited to only one bus or one type of bus. The storage medium may be implemented by any type or combination of volatile or nonvolatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.
An exemplary storage medium is coupled to the processor such the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside in an application specific integrated circuit (Application Specific Integrated Circuits, ASIC for short). It is also possible that the processor and the storage medium reside as discrete components in an electronic device or a master device.
Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the method embodiments described above may be performed by hardware associated with program instructions. The foregoing program may be stored in a computer readable storage medium. The program, when executed, performs steps including the method embodiments described above; and the aforementioned storage medium includes: various media that can store program code, such as ROM, RAM, magnetic or optical disks.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

Claims (10)

1. A credit card recommendation method, the method comprising:
acquiring demand information of a target user, historical data of a historical user and information of each credit card product;
according to the demand information of the target user, the historical data of the historical user and the information of each credit card product, determining a first recommendation list of the credit card product corresponding to the content-based recommendation algorithm, a second recommendation list of the credit card product corresponding to the user-based collaborative filtering recommendation algorithm and a third recommendation list of the credit card product corresponding to the article-based collaborative filtering recommendation algorithm through a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an article-based collaborative filtering recommendation algorithm;
and determining a target credit card product recommended for the target user through the first recommendation list, the second recommendation list and the third recommendation list based on the requirement information of the target user.
2. The method of claim 1, wherein the history data includes demand information of a history user, information of a credit card product used, and evaluation information of the credit card product; according to the demand information of the target user, the history data of the history user and the information of each credit card product, determining a first recommendation list of the credit card product corresponding to the recommendation algorithm based on the content, a second recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering, and a third recommendation list of the credit card product corresponding to the recommendation algorithm based on the collaborative filtering through a recommendation algorithm based on the content, a collaborative filtering recommendation algorithm based on the user and a collaborative filtering recommendation algorithm based on the object respectively, wherein the method comprises the following steps:
Obtaining content feature vectors of the credit card products through a TF-IDF model according to the information of the credit card products, determining recommendation lists corresponding to the credit card products through a content-based recommendation algorithm according to the content feature vectors of the credit card products, and determining a first recommendation list matched with the requirement information of the target user from the recommendation lists corresponding to the credit card products according to the requirement information of the target user and the historical data of the historical user;
determining a recommendation list corresponding to the credit card products recommended by the historical user through a collaborative filtering recommendation algorithm based on the user according to the historical data of the historical user and the information of each credit card product, and determining a second recommendation list matched with the demand information of the target user from the recommendation list corresponding to the credit card products recommended by the historical user according to the demand information of the target user and the historical data of the historical user;
according to the historical data of the historical users and the information of the credit card products, determining the similarity between the credit card products based on a collaborative filtering recommendation algorithm of the articles, constructing a recommendation list of the similar credit card products, and determining a third recommendation list matched with the requirement information of the target user from the recommendation list of the similar credit card products according to the requirement information of the target user and the historical data of the historical users.
3. The method of claim 2, wherein determining a recommendation list corresponding to each credit card product by a content-based recommendation algorithm based on the content feature vector of each credit card product comprises:
according to the content feature vector of each credit card product, obtaining the score corresponding to each credit card product through a UGC recommendation model corresponding to a content-based recommendation algorithm; the UGC recommendation model is a recommendation model for adding punishment items to hot tags;
calculating the similarity between different credit card products according to the content feature vector of each credit card product;
and determining a recommendation list corresponding to each credit card product according to the scores corresponding to each credit card product and the similarity between different credit card products, wherein the recommendation list corresponding to each credit card product comprises information of each credit card product which is sequenced based on the scores.
4. The method according to claim 2, wherein the determining, according to the historical data of the historical user and the information of each credit card product, a recommendation list corresponding to the credit card products recommended to the historical user through a collaborative filtering recommendation algorithm based on the user includes:
According to the historical data of the historical users and the information of each credit card product, determining the similarity between the historical users through a collaborative filtering recommendation algorithm based on the users;
and determining a recommendation list corresponding to the credit card products recommended by the historical users according to the similarity among the historical users, the information of the credit card products used in the historical data of the historical users and the evaluation information of the credit card products.
5. The method of claim 4, wherein said building a recommendation list for similar credit card products comprises:
according to the historical data of the historical users and the information of each credit card product, scoring data in a preset time period is obtained from a Redis cache database, wherein the scoring data are used for representing scoring data of the users on the credit card products;
constructing a recommendation list for similar credit card products according to the scoring data of each credit card product; and the recommendation list of each credit card product comprises information of other credit card products similar to the credit card product after being ranked based on the scoring data.
6. The method according to any one of claims 1-5, wherein determining a target credit card product recommended for the target user through the first recommendation list, the second recommendation list, and the third recommendation list based on the demand information of the target user, comprises:
According to the demand information of the target user, determining weights respectively corresponding to the first recommendation list, the second recommendation list and the third recommendation list in the current service scene;
determining a target credit card product recommended for the target user according to weights respectively corresponding to the first recommendation list, the second recommendation list and the third recommendation list;
correspondingly, the method further comprises the steps of:
acquiring feedback information of the target user on the recommended target credit card product; the feedback information at least comprises scoring data of the recommended result;
and optimizing a recommendation model according to the feedback information, wherein the recommendation model is determined by at least two of a content-based recommendation algorithm, a user-based collaborative filtering recommendation algorithm and an item-based collaborative filtering recommendation algorithm.
7. A credit card recommendation device, the device comprising:
the acquisition module is used for acquiring the demand information of the target user, the historical data of the historical user and the information of each credit card product;
the recommendation module is used for determining a first recommendation list of credit card products corresponding to the recommendation algorithm based on the content, a second recommendation list of credit card products corresponding to the collaborative filtering recommendation algorithm based on the user and a third recommendation list of credit card products corresponding to the collaborative filtering recommendation algorithm based on the article according to the demand information of the target user, the historical data of the historical user and the information of each credit card product through the recommendation algorithm based on the content, the collaborative filtering recommendation algorithm based on the user and the collaborative filtering recommendation algorithm based on the article;
And the recommending module is also used for determining a target credit card product recommended for the target user through the first recommending list, the second recommending list and the third recommending list based on the requirement information of the target user.
8. An electronic device, comprising: at least one processor and memory;
the memory stores computer-executable instructions;
the at least one processor executing computer-executable instructions stored in the memory causes the at least one processor to perform the credit card recommendation method of any one of claims 1 to 6.
9. A computer readable storage medium having stored therein computer executable instructions which, when executed by a processor, implement the credit card recommendation method of any one of claims 1 to 6.
10. A computer program product comprising a computer program which, when executed by a processor, implements a credit card recommendation method as claimed in any one of claims 1 to 6.
CN202311651204.7A 2023-12-04 2023-12-04 Credit card recommendation method, apparatus, device, storage medium and program product Pending CN117593089A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311651204.7A CN117593089A (en) 2023-12-04 2023-12-04 Credit card recommendation method, apparatus, device, storage medium and program product

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311651204.7A CN117593089A (en) 2023-12-04 2023-12-04 Credit card recommendation method, apparatus, device, storage medium and program product

Publications (1)

Publication Number Publication Date
CN117593089A true CN117593089A (en) 2024-02-23

Family

ID=89911381

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311651204.7A Pending CN117593089A (en) 2023-12-04 2023-12-04 Credit card recommendation method, apparatus, device, storage medium and program product

Country Status (1)

Country Link
CN (1) CN117593089A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117830046A (en) * 2024-03-06 2024-04-05 长春电子科技学院 Online course data cloud management system based on Internet

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117830046A (en) * 2024-03-06 2024-04-05 长春电子科技学院 Online course data cloud management system based on Internet
CN117830046B (en) * 2024-03-06 2024-05-07 长春电子科技学院 Online course data cloud management system based on Internet

Similar Documents

Publication Publication Date Title
CN105224699B (en) News recommendation method and device
Chen et al. AR-miner: mining informative reviews for developers from mobile app marketplace
CN108885624B (en) Information recommendation system and method
CN104978314B (en) Media content recommendations method and device
CN110334356B (en) Article quality determining method, article screening method and corresponding device
CN111461841A (en) Article recommendation method, device, server and storage medium
CN109325146A (en) A kind of video recommendation method, device, storage medium and server
KR20210047229A (en) Recommendation System and METHOD Reflecting Purchase Criteria and Product Reviews Sentiment Analysis
CN110929166A (en) Content recommendation method, electronic device and storage medium
CN111506831A (en) Collaborative filtering recommendation module and method, electronic device and storage medium
CN112862567B (en) Method and system for recommending exhibits in online exhibition
CN101211368B (en) Method for classifying search term, device and search engine system
CN117593089A (en) Credit card recommendation method, apparatus, device, storage medium and program product
Wang et al. CROWN: a context-aware recommender for web news
Wang et al. Research on hybrid collaborative filtering recommendation algorithm based on the time effect and sentiment analysis
CN113672793A (en) Information recall method and device, electronic equipment and storage medium
CN113204953A (en) Text matching method and device based on semantic recognition and device readable storage medium
CN112182351A (en) News recommendation method and device based on multi-feature fusion
CN113961823B (en) News recommendation method, system, storage medium and equipment
CN115062151A (en) Text feature extraction method, text classification method and readable storage medium
Wei et al. Online education recommendation model based on user behavior data analysis
CN111737607B (en) Data processing method, device, electronic equipment and storage medium
Kim et al. Product recommendation system based user purchase criteria and product reviews
Li et al. Exploit latent Dirichlet allocation for collaborative filtering
CN116823410A (en) Data processing method, object processing method, recommending method and computing device

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