Product recommendation method and device based on collaborative filtering, equipment and medium

Abstract

The invention provides a collaborative filtering-based product recommendation method, which can be applied to the field of artificial intelligence or cloud computing. The method comprises the steps that under the condition that a user agrees or authorizes, behavior data of at least two users are obtained, and the behavior data are used for representing the interaction relation between each user and multiple products; constructing a user product matrix based on the behavior data; based on the user product matrix, calculating the similarity between the users, and obtaining a first target user and a second target user whose similarity is greater than a preset threshold; and calculating the favorite degree of the first target user for different products, and recommending the corresponding product to a second target user who does not use the product according to the favorite degree. The invention further provides a product recommendation device and equipment based on collaborative filtering, a storage medium and a program product.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence or cloud computing, and more specifically to a product recommendation method, device, equipment, medium and program product based on collaborative filtering. Background Art

[0002] With the rapid development of the financial market and the increasing diversification of consumer demand, credit cards, as an important payment tool, are constantly enriching and expanding in types and functions. However, the existing credit card system still has many deficiencies in product recommendation, especially the lack of personalized recommendation capabilities. Many credit card systems simply display various products to users, and users choose and decide on their own. They lack consideration of users' personal needs and consumption habits, making it difficult for users to quickly find the most suitable credit card products for themselves. Alternatively, some banks make product recommendations through simple manual evaluation of users' consumption behaviors by business personnel, but due to the lack of experience of business personnel, subjective judgments are often difficult to be comprehensive, accurate and personalized. In addition, some banks promote a certain type of credit card products through publicity, which may not be suitable for other users' consumption needs. Summary of the invention

[0003] In view of the above problems, embodiments of the present disclosure provide a product recommendation method, apparatus, device, medium and program product based on collaborative filtering.

[0004] According to a first aspect of the present disclosure, a product recommendation method based on collaborative filtering is provided, the method comprising: obtaining behavior data of at least two users with the consent or authorization of the users, the behavior data being used to characterize the interaction relationship between each of the users and a plurality of products; constructing a user-product matrix based on the behavior data; calculating the similarity between the users based on the user-product matrix, and obtaining a first target user and a second target user whose similarity is greater than a preset threshold; and calculating the first target user's preference for different products, and recommending the corresponding product to a second target user who has not used the product based on the preference.

[0005] According to an embodiment of the present disclosure, the method further includes: acquiring historical behavior information of each of the users; performing data analysis on the historical behavior information to extract user features of each of the users; and adjusting the similarity of the users based on the user features.

[0006] According to an embodiment of the present disclosure, constructing a user-product matrix based on the behavioral data includes: assigning a unique user identifier to each of the users, and assigning a unique product identifier to each of the products; determining, based on the behavioral data, a usage record of each of the products by each user; and mapping the user identifiers, the product identifiers, and the usage records to obtain the user-product matrix.

[0007] According to an embodiment of the present disclosure, the similarity between the users is calculated based on the user-product matrix, and the first target user and the second target user whose similarity is greater than a preset threshold are obtained, including: for each pair of users in the user-product matrix, the similarity is calculated using a preset similarity formula according to the usage records; the calculated similarities of each pair of users are integrated into a similarity matrix, the rows and columns of the similarity matrix are used to characterize the users, and the element values ​​are used to characterize the similarity between the corresponding two users; and based on the similarity matrix, the users are filtered to obtain the first target user and the second target user whose similarity is greater than a preset threshold.

[0008] According to an embodiment of the present disclosure, the calculating of the first target user's preference for different products and recommending the corresponding products to the second target user who has not used the products according to the preference includes: calculating the first target user's preference for each product according to the behavioral data in the user-product matrix; sorting the products according to the preference, selecting the first n products from the sorted products, and filtering the products that the second target user has used to generate a recommendation list; and recommending the products in the recommendation list to the second target user.

[0009] According to an embodiment of the present disclosure, the method further includes: acquiring update information in response to the introduction of new products and / or changes to original products; and updating the user product matrix at fixed time window intervals based on the update information.

[0010] According to an embodiment of the present disclosure, the method further includes: collecting feedback data of the second target user on the product recommended in real time, and adjusting the user-product matrix according to the feedback data.

[0011] A second aspect of the present disclosure provides a product recommendation device based on collaborative filtering, the device comprising: an acquisition module, used to acquire behavior data of at least two users with the consent or authorization of the users, the behavior data being used to characterize the interaction relationship between each of the users and a plurality of products; a construction module, used to construct a user-product matrix based on the behavior data; a matching module, used to calculate the similarity between the users based on the user-product matrix, and obtain a first target user and a second target user whose similarity is greater than a preset threshold; and a recommendation module, used to calculate the first target user's preference for different products, and recommend the corresponding product to a second target user who has not used the product according to the preference.

[0012] A third aspect of the present disclosure provides an electronic device, comprising: one or more processors; a storage device for storing one or more programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the above-mentioned product recommendation method based on collaborative filtering.

[0013] The fourth aspect of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the above-mentioned product recommendation method based on collaborative filtering when executed by a processor.

[0014] The fifth aspect of the present disclosure further provides a computer program product, including a computer program, which implements the steps of the above-mentioned product recommendation method based on collaborative filtering when executed by a processor.

[0015] In the embodiments of the present disclosure, by calculating the user's similarity based on the user's behavior data, it is possible to find user groups with similar interests and preferences, thereby determining products that are similar in user behavior, so as to recommend products that meet the preferences and needs of different users, which can not only improve user satisfaction, but also increase the exposure of products that users have not used. In addition, based on the user's historical behavior information, user features are extracted, and the user features are combined with user similarities to better capture the user's personalized needs, thereby further improving the accuracy of recommendations and effectively reducing the situation where users cannot repay due to excessive consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above contents and other purposes, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0017] Figure 1 A diagram schematically shows an application scenario of a product recommendation method, apparatus, device, medium, and program product based on collaborative filtering according to an embodiment of the present disclosure;

[0018] Figure 2A flowchart of a product recommendation method based on collaborative filtering according to an embodiment of the present disclosure is schematically shown;

[0019] Figure 3 Schematically shows a flow chart of constructing a user-product matrix according to an embodiment of the present disclosure;

[0020] Figure 4 A schematic diagram of a structure of a product recommendation device based on collaborative filtering according to an embodiment of the present disclosure is shown; and

[0021] Figure 5 A block diagram of an electronic device for a product recommendation method based on collaborative filtering according to an embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0022] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0023] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0024] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0025] When using expressions such as "at least one of A, B, and C, etc.", they should generally be interpreted according to the meaning of the expression commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0026] In the technical solution of the present invention, the user information (including but not limited to user personal information, user image information, user device information, such as location information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved are all information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data comply with relevant laws, regulations and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0027] In the scenario of using personal information for automated decision-making, the methods, devices, and systems provided by the embodiments of the present disclosure provide users with corresponding operation portals for users to choose to agree or reject the automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered. The expression "automated decision-making" here refers to the activity of automatically analyzing and evaluating a person's behavioral habits, interests and hobbies, or economic, health, credit status, etc. through computer programs, and making decisions. The expression "expert decision-making" here refers to the activity of making decisions by people who specialize in a certain field, have specialized experience, knowledge and skills, and have reached a certain level of professionalism.

[0028] Figure 1 The application scenario diagram of the product recommendation method, apparatus, device, medium and program product based on collaborative filtering according to an embodiment of the present disclosure is schematically shown.

[0029] like Figure 1 As shown, the application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is used to provide a medium for a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or optical fiber cables, etc.

[0030] The user can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only for example).

[0031] The first terminal device 101, the second terminal device 102, and the third terminal device 103 may be various electronic devices having display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0032] The server 105 may be a server that provides various services, such as a background management server (only as an example) that provides support for websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103. The background management server may analyze and process the received data such as user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

[0033] It should be noted that the product recommendation method based on collaborative filtering provided in the embodiment of the present disclosure can generally be executed by the server 105. Accordingly, the product recommendation device based on collaborative filtering provided in the embodiment of the present disclosure can generally be set in the server 105. The product recommendation method based on collaborative filtering provided in the embodiment of the present disclosure can also be executed by a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Accordingly, the product recommendation device based on collaborative filtering provided in the embodiment of the present disclosure can also be set in a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105.

[0034] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0035] The following will be based on Figure 1 The scene described by Figure 2~Figure 3 The product recommendation method based on collaborative filtering according to an embodiment of the present disclosure is described in detail.

[0036] Figure 2 The flowchart of the product recommendation method based on collaborative filtering according to an embodiment of the present disclosure is schematically shown.

[0037] like Figure 2 As shown, the product recommendation method based on collaborative filtering in this embodiment includes operations S210 to S240.

[0038] In operation S210, with the consent or authorization of the users, behavior data of at least two users are obtained, where the behavior data is used to characterize the interaction relationship between each of the users and a plurality of products.

[0039] In an embodiment of the present disclosure, before obtaining the user's behavior data, the user's consent or authorization is obtained. For example, before operation S210, a request to obtain the behavior data is issued to the user. When the user agrees or authorizes that the behavior data can be obtained, operation S210 is performed.

[0040] In operation S220, a user-product matrix is ​​constructed based on the behavior data.

[0041] In operation S230, based on the user-product matrix, the similarity between the users is calculated, and a first target user and a second target user whose similarity is greater than a preset threshold are obtained.

[0042] In operation S240, the first target user's preference for different products is calculated, and the corresponding product is recommended to a second target user who has not used the product according to the preference.

[0043] In an embodiment of the present disclosure, a corresponding operation entry is provided for the user to choose to agree or reject the automated decision result. That is, before the process / decision of recommending the corresponding product to the second target user who has not used the product according to the likeability, an instruction of the user to agree or reject the process / decision input through the corresponding operation entry is obtained. If the user agrees to the process / decision, the process / decision of recommending the corresponding product to the second target user who has not used the product according to the likeability is executed, that is, step S240 is executed. If the user refuses to perform the process / decision, the expert decision process is entered.

[0044] In some exemplary embodiments, with the consent or authorization of the user, the interaction data (i.e., behavior data) between the user and the product is collected from various data sources (such as e-commerce platforms, social media, or browser history, etc.), such as the user's consumption habits, purchase records (the most frequently used credit card products), browsing records, evaluation records, and click records, etc., to obtain the user's interest and preference for different products. Then, the collected user behavior data is organized into a matrix form, where the rows represent users and the columns represent products. The element values ​​in the matrix can represent the frequency or intensity of a certain behavior of the user on the product (such as purchase, browsing, etc.), thereby quantifying the degree of association between the user and the product. Then, an appropriate similarity calculation method (such as cosine similarity, Pearson correlation coefficient, etc.) is used to calculate the similarity between users. The larger the similarity value, the more similar the behaviors of the two users are, and thus they may have similar product preferences. Then, based on the similarity calculation results, the user pairs whose similarity is greater than the preset threshold are screened out as the first target user and the second target user. Among them, the preset threshold can be adjusted according to the actual application scenario and data characteristics to balance the accuracy and diversity of the recommendation results. Finally, for the first target user, the frequency, intensity or time factors of the historical behavior data can be weighted to calculate his preference for different products, so that the products liked by the first target user (and not used by the second target user) can be recommended to the second target user as personalized recommendation results.

[0045] It should be noted that, from the screened user combinations, one or more pairs are arbitrarily selected as the first target user and the second target user. Since the similarity is symmetrical, any one of the selected user pairs can be regarded as the first or second target user. At the same time, the first and second target users can also be regarded as user groups, that is, it can be considered that the number of users of the first and second target users is greater than or equal to 3, and it is no longer limited to the similarity between the two users, so that the product liked by a certain person in the user group can be recommended to other users in the user group who have not used the product. In addition, more complex algorithms, such as cluster analysis, matrix decomposition, etc., can also be used to calculate the similarity between user groups, and then calculate the average liking or weighted liking of one of the user groups for the product, and then recommend the product to each user in the other user group based on these likings.

[0046] It is understandable that by calculating the similarity of users based on their behavioral data, user characteristics are combined with products, and similar products are recommended between similar users, so as to accurately capture user preferences and needs, realize personalized product recommendations, and improve user satisfaction. In addition, recommending products to target users who have not yet used similar products can effectively increase product exposure and market share, and help expand users' understanding and selection of different products.

[0047] Figure 3 The flowchart of constructing a user-product matrix according to an embodiment of the present disclosure is schematically shown.

[0048] In some exemplary embodiments, Figure 3 As shown, constructing a user-product matrix based on the behavioral data includes: assigning a unique user identifier to each user, and assigning a unique product identifier to each product; determining the usage record of each user for each product based on the behavioral data; and mapping the user identifier, the product identifier and the usage record to obtain the user-product matrix.

[0049] In some exemplary embodiments, in order to ensure that each user and product can be uniquely and clearly identified in the user-product matrix, a user identifier can be assigned to each user, and a product identifier can also be assigned to each product. Then, the behavioral data is analyzed in depth to determine and record the specific usage records of each user for each product. The usage records may include multiple dimensions such as the number of times the user browses the product, purchase history, usage duration, and evaluation feedback. Finally, the user identifier, product identifier, and usage record are mapped to obtain the user-product matrix.

[0050] For example, suppose there are three users (user identifiers are A, B, C) and four credit card products (product identifiers are X, Y, Z). Then, for each user and product, create a record to record the number of times the user uses the product. Then, a user-product matrix can be constructed, where each row of the matrix represents a user, each column represents a credit card product, and the matrix elements represent the number of times the user uses the credit card product.

[0051] Table 1

[0052]

[0053] It is understandable that through the user-product matrix, we can clearly see each user's usage and preferences for each product, which helps to gain a deeper understanding of user behavior patterns and thus build an accurate user portrait. At the same time, based on the behavioral data in the user-product matrix, we can provide users with personalized product recommendations.

[0054] On the basis of the above-mentioned embodiments, in this embodiment, the similarity between the users is calculated based on the user-product matrix, and the first target user and the second target user whose similarity is greater than a preset threshold are obtained, including: for each pair of users in the user-product matrix, the similarity is calculated using a preset similarity formula according to the usage records; the calculated similarities of each pair of users are integrated into a similarity matrix, the rows and columns of the similarity matrix are used to characterize the users, and the element values ​​are used to characterize the similarity between the corresponding two users; and based on the similarity matrix, the users are filtered to obtain the first target user and the second target user whose similarity is greater than a preset threshold.

[0055] Optionally, cosine similarity is used as the calculation standard, and its value is between -1 and 1, where 1 means complete similarity (same direction), -1 means complete opposite (opposite direction), and 0 means no linear correlation (orthogonal). The formula for cosine similarity is:

[0056]

[0057] Among them, A and B represent the usage times vectors of two users, n is the number of products, A i and B i Represent the number of times user A and user B use the i-th product.

[0058] All the calculated user similarity values ​​are integrated into a matrix, where the rows and columns of the matrix represent users, and the element values ​​represent the similarity between the corresponding two users. According to the preset similarity threshold, all user combinations with similarities exceeding this threshold are screened out. For example, according to the calculated cosine similarity value, if the similarity value of user A and user B is close to 1, then it can be considered that they are very similar in their credit card product usage habits, and user A and user B can be used as the first target user and the second target user.

[0059] For example, the similarity can be directly calculated based on the user's preference for different products. For example, the similarity can be calculated based on the number of times the user uses different credit cards to consume products, that is, the similar first target user and second target user are determined based on the frequency of use. In the above embodiment, user A and user B have a high frequency of use of product X, so they can be considered to be the first target user and the second target user with high similarity.

[0060] It is understandable that by calculating the similarity between users, it is possible to find user groups with similar interests and preferences, thereby better understanding the characteristics and needs of user groups, determining credit card products that are similar in user behavior, and making targeted recommendations of similar or related products, ultimately improving the accuracy of recommendations and user satisfaction.

[0061] In an embodiment of the present disclosure, the method further includes: acquiring historical behavior information of each of the users; performing data analysis on the historical behavior information to extract user features of each of the users; and adjusting the similarity of the users based on the user features.

[0062] In some exemplary embodiments, historical behavior information of users can be obtained from various channels, including but not limited to the user's credit history, transaction records, historical overdraft amount, debit card usage information and user assets, etc., to build a more in-depth and comprehensive user portrait. Then, data analysis is performed on the historical behavior information to extract the user's characteristics, which can not only include the user's basic attributes (such as age, gender, geographic location, etc.), but also involve the user's interest preferences, consumption level, shopping habits, payment mode, and social influence. Furthermore, the similarity between users can be adjusted according to user characteristics to combine user characteristics with user similarity, so as to better capture the user's personalized needs.

[0063] It should be noted that in the process of extracting user characteristics, we can focus on analyzing the user's financial data. For example, we can deeply analyze the user's credit history to assess his credit rating and default risk. At the same time, we can also carefully study the user's transaction records and overdraft amounts to determine his spending power and payment habits. In addition, through the analysis of debit card usage information and user asset data, we can have a more comprehensive understanding of the user's financial status and investment preferences.

[0064] It should be noted that when conducting data analysis, strict data encryption and anonymization processing technology is used to ensure the absolute security and privacy protection of user information. At the same time, a complete data use and management system can be established to regulate all aspects of data collection, storage, processing and use.

[0065] It can be understood that when determining the similarity of users, the embodiments of the present disclosure can not only determine it based on the interactive relationship between the user and the product, but also based on the user's characteristics (such as the user's credit and consumption situation), so as to personalize product recommendations based on the user's situation, maximize the advantages of each product, and effectively reduce the situation where users are unable to repay due to excessive consumption.

[0066] On the basis of the above-mentioned embodiments, in this embodiment, the calculation of the first target user's preference for different products and the recommendation of the corresponding products to the second target user who has not used the product according to the preference includes: calculating the first target user's preference for each product according to the behavior data in the user-product matrix; sorting the products according to the preference, selecting the first n products from the sorted products, and filtering the products that the second target user has used to generate a recommendation list; and recommending the products in the recommendation list to the second target user.

[0067] According to the behavior data in the user-product matrix, the first target user (i.e., the identified user with similar interests or needs) can calculate the preference for each product. In this process, one or more factors may be involved, such as the frequency of user purchases, browsing time, evaluation level, and number of shares. Then, the above factors can be converted into quantitative indicators in a variety of ways, such as weighted average, machine learning models (such as random forests, gradient boosting trees, etc.) or deep learning models (such as neural networks) to obtain the user's preference for the product. After calculating the first target user's preference for all products, they can be sorted according to the size of the preference to identify the user's favorite products. Then, the first n (n is the preset number of recommendations) of the sorted products are selected as potential recommended products. At the same time, in order to avoid recommending products that the second target user has used, the recommendation list can be filtered to remove products that have appeared in the second target user's historical records. Finally, a personalized recommendation list for the second target user can be generated.

[0068] For example, referring to Table 1, for the first target user A, the credit card products frequently used by user A are identified, and based on these products, other users (i.e., the second target user B) who also frequently use the products are found. Then, in addition to the products that user B has used, other products (such as product Z) that user B has frequently used are checked, and user A has not used product Z, so product Z can be recommended to user A. Similarly, product Y can be recommended to user B who is similar to user A.

[0069] It can be understood that by calculating the likeability and making recommendations based on the likeability, it is ensured that the recommended products are not only highly matched with the user's interests, but also directly reflect the user's real needs and preferences, thereby enabling more accurate recommendations.

[0070] In an embodiment of the present disclosure, the method further includes: acquiring update information in response to the introduction of new products and / or changes in original products; and updating the user product matrix at fixed time window intervals based on the update information.

[0071] When new credit card products are introduced to the market or existing products are changed (such as function upgrades, rate adjustments, etc.), update information can be obtained from the product management department, marketing department or third-party data sources through database queries, API interface calls and file imports. Based on the updated information obtained, the user product matrix can be updated at fixed time windows (such as daily, weekly or monthly), for example, it can include operations such as adding new products, modifying original products, and deleting outdated products. When updating the user product matrix, you can also re-evaluate the relationship between users and products to ensure that the matrix can accurately reflect the latest status between users and products.

[0072] In order to achieve real-time recommendation, the product recommendation method provided by the embodiment of the present disclosure can be integrated into a real-time system. In the real-time system, the recommendation method can be made to respond quickly to data changes, such as product updates, user updates, or behavior data updates, by optimizing the algorithm model, such as using incremental learning or online learning technology, so as to update the recommendation results in real time.

[0073] It is understandable that the user product matrix is ​​regularly updated with updated product information, thereby reflecting changes between products and the introduction of new products in real time, so that the latest and most accurate product recommendations can be provided to users.

[0074] In an embodiment of the present disclosure, the method further includes: collecting feedback data of the second target user on the product recommended in real time, and adjusting the user-product matrix according to the feedback data.

[0075] After making recommendations to the second target user based on the recommendation list, you can also collect user feedback data on the real-time recommended products, such as the user's click-through rate, purchase behavior, ratings, comments, and other forms of interaction records for the recommended products. Then, you can clean, organize, and analyze the collected feedback data to extract valuable target information. Finally, based on the analysis results of the feedback data, you can make corresponding adjustments to the user-product matrix, such as updating the user's preference for the product, introducing new user features or product attributes, adjusting the similarity calculation method or weight, and optimizing the logic and parameters of the recommendation algorithm, so that the matrix can more accurately reflect the user's current preferences and market dynamics.

[0076] The user-product matrix is ​​not only a static data structure, but also a dynamic intelligent system that can continuously evolve as user behavior data is updated. By real-time monitoring and analyzing the latest changes in user behavior data, the elements in the matrix can be adjusted in a timely manner to ensure that the matrix can always accurately reflect the latest relationship between users and products.

[0077] It is understandable that by collecting user feedback, we can discover users' new demands and preferences for products and changes in market trends, so as to adjust the user-product matrix so that the matrix can more accurately reflect the relationship between users and products, thereby further improving the accuracy of recommendations.

[0078] Figure 4 The structural block diagram of a product recommendation device based on collaborative filtering according to an embodiment of the present disclosure is schematically shown.

[0079] like Figure 4 As shown, the product recommendation device 400 based on collaborative filtering according to this embodiment includes an acquisition module 410 , a construction module 420 , a matching module 430 and a recommendation module 440 .

[0080] The acquisition module 410 is used to acquire the behavior data of at least two users with the consent or authorization of the users, and the behavior data is used to characterize the interaction relationship between each of the users and multiple products. In one embodiment, the acquisition module 410 can be used to perform the operation S210 described above, which will not be repeated here.

[0081] The construction module 420 is used to construct a user-product matrix based on the behavior data. In one embodiment, the construction module 420 can be used to perform the operation S220 described above, which will not be described in detail here.

[0082] The matching module 430 is used to calculate the first target user's preference for different products, and recommend the corresponding product to the second target user who has not used the product according to the preference. In one embodiment, the matching module 430 can be used to perform the operation S230 described above, which will not be described in detail here.

[0083] The recommendation module 440 is used to predict the degree of change of the performance data in the future time period based on the performance characteristics and using the regression prediction layer of the regression prediction model to warn the target database. In one embodiment, the recommendation module 440 can be used to perform the operation S240 described above, which will not be repeated here.

[0084] In the embodiment of the present disclosure, the matching module 430 may also be used to: obtain historical behavior information of each of the users; perform data analysis on the historical behavior information to extract user features of each of the users; and adjust the similarity of the users based on the user features.

[0085] In an embodiment of the present disclosure, the construction module 420 is specifically used to: assign a unique user identifier to each of the users, and assign a unique product identifier to each of the products; determine the usage record of each of the users for each of the products based on the behavioral data; and map the user identifier, the product identifier and the usage record to obtain the user-product matrix.

[0086] In an embodiment of the present disclosure, the matching module 430 is specifically used to: for each pair of users in the user-product matrix, calculate the similarity using a preset similarity formula according to the usage record; integrate the calculated similarities of each pair of users into a similarity matrix, the rows and columns of the similarity matrix are used to characterize the users, and the element values ​​are used to characterize the similarities between the corresponding two users; and based on the similarity matrix, filter the users to obtain the first target user and the second target user whose similarity is greater than a preset threshold.

[0087] In an embodiment of the present disclosure, the recommendation module 440 is specifically used to: calculate the first target user's preference for each product based on the behavioral data in the user-product matrix; sort the products according to the preference, select the first n products from the sorted products, and filter the products that the second target user has used to generate a recommendation list; and recommend the products in the recommendation list to the second target user.

[0088] In an embodiment of the present disclosure, the construction module 420 may also be used to: obtain update information in response to the introduction of new products and / or changes to original products; and update the user-product matrix at fixed time windows based on the update information.

[0089] In an embodiment of the present disclosure, the recommendation module 440 may also be used to: collect feedback data of the second target user on the product recommended in real time, and adjust the user-product matrix according to the feedback data.

[0090] According to an embodiment of the present disclosure, any multiple modules among the acquisition module 410, the construction module 420, the matching module 430 and the recommendation module 440 can be combined into one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the acquisition module 410, the construction module 420, the matching module 430 and the recommendation module 440 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation modes of software, hardware and firmware or in any appropriate combination of any of them. Alternatively, at least one of the acquisition module 410, the construction module 420, the matching module 430 and the recommendation module 440 can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be performed.

[0091] Figure 5 A block diagram of an electronic device for a product recommendation method based on collaborative filtering according to an embodiment of the present disclosure is schematically shown.

[0092] like Figure 5 As shown, the electronic device 500 according to an embodiment of the present invention includes a processor 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage part 508 to a random access memory (RAM) 503. The processor 501 may include, for example, a general-purpose microprocessor (such as a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (for example, an application-specific integrated circuit (ASIC)), etc. The processor 501 may also include an onboard memory for caching purposes. The processor 501 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present invention.

[0093] In RAM 503, various programs and data required for the operation of electronic device 500 are stored. Processor 501, ROM 502 and RAM 503 are connected to each other via bus 504. Processor 501 performs various operations of the method flow according to the embodiment of the present invention by executing the program in ROM 502 and / or RAM 503. It should be noted that the program can also be stored in one or more memories other than ROM 502 and RAM 503. Processor 501 can also perform various operations of the method flow according to the embodiment of the present invention by executing the program stored in the one or more memories.

[0094] According to an embodiment of the present invention, the electronic device 500 may further include an input / output (I / O) interface 505, which is also connected to the bus 504. The electronic device 500 may further include one or more of the following components connected to the input / output (I / O) interface 505: an input portion 506 including a keyboard, a mouse, etc.; an output portion 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage portion 508 including a hard disk, etc.; and a communication portion 509 including a network interface card such as a LAN card, a modem, etc. The communication portion 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the input / output (I / O) interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed, so that the computer program read therefrom is installed into the storage portion 508 as needed.

[0095] The present invention also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiment; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present invention is implemented.

[0096] According to an embodiment of the present invention, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, an apparatus or a device. For example, according to an embodiment of the present invention, the computer-readable storage medium may include the ROM 502 and / or RAM 503 described above and / or one or more memories other than ROM 502 and RAM 503.

[0097] The embodiment of the present disclosure also includes a computer program product, which includes a computer program, and the computer program contains program code for executing the method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to enable the computer system to implement the product recommendation method based on collaborative filtering provided by the embodiment of the present disclosure.

[0098] The above functions defined in the system / device of the embodiment of the present disclosure are performed when the computer program is executed by the processor 501. According to the embodiment of the present disclosure, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0099] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices, magnetic storage devices, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of signals on a network medium, and downloaded and installed through the communication part 509, and / or installed from the removable medium 511. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0100] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 509, and / or installed from the removable medium 511. When the computer program is executed by the processor 501, the above functions defined in the system of the embodiment of the present disclosure are performed. According to the embodiment of the present disclosure, the system, device, means, module, unit, etc. described above can be implemented by a computer program module.

[0101] According to an embodiment of the present disclosure, the program code for executing the computer program provided by the embodiment of the present disclosure can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, Java, C++, python, "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).

[0102] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0103] It will be appreciated by those skilled in the art that the features described in the various embodiments of the present disclosure may be combined and / or combined in a variety of ways, even if such combinations or combinations are not explicitly described in the present disclosure. In particular, without departing from the spirit and teachings of the present disclosure, the features described in the various embodiments of the present disclosure may be combined and / or combined in a variety of ways. All of these combinations and / or combinations fall within the scope of the present disclosure.

[0104] The embodiments of the present disclosure are described above. However, these embodiments are only for illustrative purposes and are not intended to limit the scope of the present disclosure. Although the embodiments are described above, this does not mean that the measures in the various embodiments cannot be used in combination to advantage. Without departing from the scope of the present disclosure, those skilled in the art may make a variety of substitutions and modifications, which should all fall within the scope of the present disclosure.

Claims

1. A product recommendation method based on collaborative filtering, characterized in that: The method comprises: With the consent or authorization of the users, obtaining behavior data of at least two users, wherein the behavior data is used to characterize the interaction relationship between each of the users and the plurality of products; Based on the behavioral data, a user-product matrix is ​​constructed; Based on the user-product matrix, calculating the similarity between the users, and obtaining a first target user and a second target user whose similarity is greater than a preset threshold; and The first target user's preference for different products is calculated, and the corresponding product is recommended to a second target user who has not used the product according to the preference.

2. The method according to claim 1, characterized in that The method further comprises: Obtaining historical behavior information of each of the users; Performing data analysis on the historical behavior information to extract user characteristics of each of the users; and Based on the user characteristics, the similarity of the users is adjusted.

3. The method according to claim 1 or 2, characterized in that: The constructing of a user-product matrix based on the behavior data includes: assigning a unique user identifier to each of the users, and assigning a unique product identifier to each of the products; Based on the behavior data, determining a usage record of each of the products by each of the users; and The user identifier, the product identifier and the usage record are mapped to obtain the user-product matrix.

4. The method according to claim 3, characterized in that The calculating the similarity between the users based on the user-product matrix and obtaining the first target user and the second target user whose similarity is greater than a preset threshold comprises: For each pair of users in the user-product matrix, the similarity is calculated using a preset similarity formula according to the usage records; Integrate the calculated similarities of each pair of the users into a similarity matrix, wherein the rows and columns of the similarity matrix are used to represent the users, and the element values ​​are used to represent the similarities between the corresponding two users; and Based on the similarity matrix, the users are filtered to obtain a first target user and a second target user whose similarity is greater than a preset threshold.

5. The method according to claim 1, 2 or 4, characterized in that: The calculating the first target user's preference for different products and recommending the corresponding product to a second target user who has not used the product according to the preference comprises: Calculating the first target user's preference for each product based on the behavior data in the user-product matrix; Sort the products according to the degree of preference, select the first n products from the sorted products, and filter the products that the second target user has used to generate a recommendation list; and Recommend the product in the recommendation list to the second target user.

6. The method according to claim 1 or 2, characterized in that: The method further comprises: Obtaining updated information in response to the introduction of new products and / or changes to original products; and Based on the update information, the user-product matrix is ​​updated at fixed time window intervals.

7. The method according to claim 6, characterized in that The method further comprises: Collect feedback data of the second target user on the product recommended in real time, and adjust the user-product matrix according to the feedback data.

8. A product recommendation device based on collaborative filtering, characterized in that: The device comprises: An acquisition module, used to acquire behavior data of at least two users with the consent or authorization of the users, wherein the behavior data is used to characterize the interaction relationship between each of the users and the plurality of products; A construction module, used to construct a user-product matrix based on the behavior data; A matching module, configured to calculate the similarity between the users based on the user-product matrix, and obtain a first target user and a second target user whose similarity is greater than a preset threshold; and The recommendation module is used to calculate the first target user's preference for different products, and recommend the corresponding product to the second target user who has not used the product according to the preference.

9. An electronic device, comprising: one or more processors; a storage device for storing one or more computer programs, It is characterized in that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

11. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images