Information generation method and device, equipment, storage medium and program product

Abstract

The invention provides an information generation method and device, equipment, a storage medium and a program product, and can be applied to the technical field of big data. The information generation method comprises the following steps: obtaining a first historical purchase record that a plurality of historical users have purchased a plurality of products, the first historical purchase record comprising transaction information corresponding to each of the plurality of products; performing data statistical processing on the basis of the transaction information corresponding to the plurality of products to obtain respective purchase frequencies of the plurality of products; determining a target product combination according to the respective purchase frequency of the plurality of products; and generating product recommendation information for the plurality of target users according to the second historical purchase records of the plurality of target users and the product information of the target product combination.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of big data, in particular to an information generation method and device, equipment, a storage medium and a program product. BACKGROUND

[0002] In the product recommendation technical field, a collaborative filtering algorithm based on user preferences is often used. This method realizes recommendation by mining the similarity between user historical purchase behavior and group behavior. However, the collaborative filtering algorithm has a cold start problem, that is, new users have no historical interaction data, new items are not contacted by enough users, or the overall interaction data is scarce when the system is initially built, which leads to inaccurate recommendation. At the same time, the sparse matrix problem is caused by the large number of users and items, and each user only interacts with a small number of items, so that most elements in the interaction matrix are zero, resulting in extremely sparse data and affecting similarity calculation and recommendation effect. SUMMARY

[0003] In view of the above problems, the present application provides an information generation method, device, equipment, medium and program product.

[0004] According to the first aspect of the present application, an information generation method is provided, comprising: according to the embodiments of the present application, obtaining a plurality of historical user first historical purchase records of a plurality of products, wherein the first historical purchase record comprises transaction information corresponding to each of the plurality of products; based on the transaction information corresponding to each of the plurality of products, performing data statistical processing to obtain the frequency of purchase of each of the plurality of products; determining a target product combination according to the frequency of purchase of each of the plurality of products; and generating product recommendation information for a plurality of target users according to the second historical purchase records of the plurality of target users and the product information of the target product combination.

[0005] According to the embodiments of the present application, the information generation method further comprises: obtaining first attribute information of a plurality of historical users and second attribute information of a plurality of target users; classifying the plurality of historical users according to the first attribute information to obtain a plurality of historical user groups; and classifying the plurality of target users according to the second attribute information to obtain a plurality of target user groups.

[0006] According to the embodiments of the present application, the product recommendation information for the plurality of target users is generated according to the second historical purchase records of the plurality of target users and the product information of the target product combination, comprising: in the case that at least one target user group is the same as at least one historical user group, generating product recommendation information for at least one target user according to the second historical purchase records of the at least one target user and the product information of the target product combination corresponding to the at least one historical user group.

[0007] According to an embodiment of the present application, the transaction information comprises a transaction number, and the data statistical processing is performed based on the transaction information corresponding to each of the plurality of products to obtain the frequency of purchase of each of the plurality of products, comprising: calculating the frequency of purchase of each of the plurality of products in the plurality of transaction numbers to obtain the frequency of purchase of each of the plurality of products.

[0008] According to an embodiment of the present application, the target product combination is determined according to the frequency of purchase of each of the plurality of products, comprising: determining a plurality of preliminary products according to the frequency of purchase of each of the plurality of products and a first frequency threshold; obtaining a plurality of candidate product combinations according to the plurality of preliminary products; and determining at least one target product combination according to the plurality of candidate product combinations and a second frequency threshold.

[0009] According to an embodiment of the present application, the plurality of candidate product combinations is obtained according to the plurality of preliminary products, comprising: combining the plurality of preliminary products two by two to obtain a plurality of preliminary product combinations; and combining the preliminary product combinations satisfying the second frequency threshold two by two, and so on until no combination satisfying the second frequency threshold appears, to obtain the plurality of candidate product combinations.

[0010] According to an embodiment of the present application, the product recommendation information for the plurality of target users is generated according to the second historical purchase records of the plurality of target users and the product information of the target product combination, comprising: determining a plurality of products purchased by a target user according to the second historical purchase records; determining at least one reference product combination from the target product combination according to the plurality of products purchased by the target user; determining products not purchased by the target user according to the product information of the reference product combination; and generating product recommendation information for the target user according to the products not purchased by the target user.

[0011] The second aspect of the present application provides an information generation device, comprising: a first acquisition module configured to acquire first historical purchase records of a plurality of products purchased by a plurality of historical users, wherein the first historical purchase records comprise transaction information corresponding to each of the plurality of products; a statistical module configured to perform data statistical processing based on the transaction information corresponding to each of the plurality of products to obtain the frequency of purchase of each of the plurality of products; a determination module configured to determine a target product combination according to the frequency of purchase of each of the plurality of products; and a generation module configured to generate product recommendation information for a plurality of target users according to second historical purchase records of the plurality of target users and product information of the target product combination.

[0012] The third aspect of the present application provides an electronic device, comprising: one or more processors; a memory configured to store one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the above method.

[0013] The fourth aspect of the present application further provides a computer readable storage medium, having stored thereon a computer program or instructions, which, when executed by a processor, implement the steps of the above method.

[0014] The fifth aspect of the present application further provides a computer program product, comprising a computer program or instructions, which, when executed by a processor, implement the steps of the above method.

[0015] According to embodiments of the present application, the target product combination is obtained according to the purchase frequency of the historical users, i.e., the correlation between products and products is obtained, so that even if the historical data of the target user is limited, e.g., the target user has only purchased one product, the target user can still be recommended according to the historical user preferences, to a certain extent, avoiding the problem of sparse matrix; on the other hand, the present application only needs to scan the transaction data (the first historical purchase record) once to complete the statistics of all product purchase frequencies, compared with the traditional algorithm of generating frequent item sets, repeatedly scanning, reducing I / O operation and memory occupation, thereby reducing the system resource consumption. BRIEF DESCRIPTION OF DRAWINGS

[0016] The above and other objects, features and advantages of the present application will become more apparent from the following description of embodiments of the present application, taken in conjunction with the accompanying drawings, in which:

[0017] Figure 1 An application scenario diagram of the information generation method, device, equipment, medium and program product according to embodiments of the present application is schematically shown;

[0018] Figure 2 A flowchart of the information generation method according to embodiments of the present application is schematically shown;

[0019] Figure 3 A flowchart of the method for determining the target product combination according to embodiments of the present application is schematically shown;

[0020] Figure 4 A flowchart of the method for determining the candidate product combination according to embodiments of the present application is schematically shown;

[0021] Figure 5 A flowchart of the method for determining the product recommendation information according to embodiments of the present application is schematically shown;

[0022] Figure 6 A structural block diagram of the information generation device according to embodiments of the present application is schematically shown; and

[0023] Figure 7 A block diagram of an electronic device suitable for implementing the information generation method according to embodiments of the present application is schematically shown. DETAILED DESCRIPTION

[0024] Hereinafter, embodiments of the present application will be described with reference to the accompanying drawings. It is to be understood, however, that these descriptions are merely exemplary and are intended to illustrate the scope of the present application, not to limit it. In the following detailed description of the embodiments, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the present application. However, it would be apparent to those skilled in the art that the embodiments, or portions thereof, can be practiced without such specific details. In other instances, well-known structures and

[0025] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used herein, the term "comprises", "comprising", or the like, means the presence of stated features, steps, operations, and / or components, but does not preclude the presence or addition of one or more other features, steps, operations, or components.

[0026] All terms used herein, including technical and scientific terms, have the meanings commonly understood by one of ordinary skill in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having meanings that are consistent with the context of the specification, and should not be interpreted in an idealized or overly formal manner.

[0027] In the case of using expressions similar to "at least one of A, B, and C, etc.", it should generally be interpreted to include any of them alone, any combination of two or more of them, and the like in a manner that makes sense in the context of the specification.

[0028] It should be noted that the information generation method and device of the present application can be used in the field of financial technology and the field of big data technology, and can also be used in any field other than the field of financial technology and the field of big data. The application field of the information generation method and device of the present application is not limited.

[0029] In the technical solutions of the present application, 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 for analysis, stored data, displayed data, etc.) involved are information and data authorized by the user or authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards, necessary security measures are taken, do not violate public order and good customs, and provide corresponding operation portal for user to choose authorization or refusal.

[0030] In the scenario of making automated decisions by using personal information, the method, device and system provided by the embodiments of the present application all provide corresponding operation entrances for the user to select to agree or reject the automated decision result; if the user selects to reject, the expert decision process is entered. The expression "automated decision" herein refers to the activity of making decisions by automatically analyzing and evaluating the behavior habits, interests and hobbies, or economic, health and credit conditions of a person by using a computer program. The expression "expert decision" herein refers to the activity of making decisions by a person who is specialized in a certain field, has certain professional level of experience, knowledge and skills.

[0031] The Apriori algorithm needs to scan the transaction database multiple times when calculating the support degree of an item set, such as scanning once to generate a frequent 1-item set and repeatedly scanning to generate a frequent multi-item set, which is low in efficiency. Meanwhile, a large number of candidate sets are generated in the process of generating a frequent item set, and the number of candidate sets increases exponentially with the increase of the number of item sets, consuming a large amount of computing resources and time.

[0032] The Fp-Growth algorithm mines frequent item sets by constructing an FP tree, but when processing sparse transaction data, the FP tree is large in size and occupies a large amount of memory, affecting the performance and scalability of the algorithm.

[0033] Therefore, the embodiments of the present application provide an information generation method: obtaining a first historical purchase record of a plurality of products purchased by a plurality of historical users, wherein the first historical purchase record comprises transaction information corresponding to each of the plurality of products; performing data statistical processing based on the transaction information corresponding to each of the plurality of products to obtain the frequency of purchase of each of the plurality of products; determining a target product combination according to the frequency of purchase of each of the plurality of products; and generating product recommendation information for a plurality of target users according to the second historical purchase record of the plurality of target users and the product information of the target product combination.

[0034] Figure 1 An application scenario diagram of the information generation method according to the embodiments of the present application is schematically shown.

[0035] As shown in Figure 1 The application scenario 100 according to the embodiments can 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 as a medium to provide 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 can include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0036] A user can use at least one of 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 as examples).

[0037] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smartphones, tablet computers, laptop computers, desktop computers, etc.

[0038] The server 105 can be a server providing various services, such as a background management server providing support for websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (only as examples). The background management server can analyze and process received user requests and other data, and feed back the processing results (such as web pages, information, or data, etc. obtained or generated according to user requests) to the terminal devices.

[0039] It should be noted that the information generation method provided by the embodiments of the present disclosure can generally be executed by the server 105. Correspondingly, the information generation apparatus provided by the embodiments of the present disclosure can generally be arranged in the server 105. The information generation method provided by the embodiments of the present disclosure can also be executed by a server or a server cluster different from the server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or the server 105. Correspondingly, the information generation apparatus provided by the embodiments of the present disclosure can also be arranged in a server or a server cluster different from the server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or the server 105.

[0040] It should be understood that Figure 1 The number of terminal devices, networks, and servers in the above description is only illustrative. According to the needs of implementation, there can be any number of terminal devices, networks, and servers.

[0041] The information generation method according to the embodiments of the present disclosure will be described in detail below based on the scenarios described above. Figure 1 Figures 2-5 The information generation method according to the embodiments of the present disclosure will be described in detail below based on the scenarios described above.

[0042] Figure 2 The flowchart of the information generation method according to the embodiments of the present disclosure is schematically shown.

[0043] As​Figure 2 As shown, the information generation method 200 of this embodiment includes operation S210 to operation S240.

[0044] In operation S210, a plurality of first historical purchase records of a plurality of products purchased by a plurality of historical users are acquired, wherein the first historical purchase records include transaction information corresponding to the plurality of products respectively.

[0045] According to an embodiment of the present application, the historical user can be a user who has purchased a related product within a predetermined historical time range, and the first historical purchase record is a record of different products purchased by a plurality of historical users, each record containing transaction information of the corresponding product, which can be a transaction number, a transaction time, a transaction amount, a user identifier, etc. The predetermined historical time range can be one week, one month, one quarter, one year, etc., which is not limited herein.

[0046] In operation S220, data statistical processing is performed based on the transaction information corresponding to the plurality of products respectively, to obtain a frequency at which the plurality of products are purchased respectively.

[0047] According to an embodiment of the present application, the transaction information of each product is processed to calculate the number of times each product is purchased by a historical user, reflecting the popularity of the product.

[0048] In operation S230, a target product combination is determined according to the frequency at which the plurality of products are purchased respectively.

[0049] According to an embodiment of the present application, a product combination with high frequency of purchase is selected according to the frequency of purchase of the product, wherein the product combination is used to reflect the frequency at which the plurality of products are purchased together, i.e., the co-occurrence frequency. If the frequency of purchase of a product is low, then the frequency of purchase of the product together with other products must also be low. If the frequency of purchase of a product and another product is high, then the frequency of purchase of the product alone must also be high. Therefore, the product combination can reflect the association between the products to some extent, for example, if the target product combination is

product A, product B

[0050] In operation S240, product recommendation information for a plurality of target users is generated according to second historical purchase records of the plurality of target users and product information of the target product combination.

[0051] According to an embodiment of the present application, the target user can be any one of the historical users described above, or can be unrelated to the historical users described above. For a user who has purchased any product, personalized product recommendation content can be generated for the target user by the products purchased by the user and the target product combination obtained through the above process.

[0052] According to the embodiments of the present application, the target product combination is obtained according to the purchase frequency of the historical users, that is, the correlation between the products and the products purchased is obtained, so that even if the historical data of the target user is limited, for example, the target user has only purchased one product, the recommendation can still be made according to the historical user preferences, and to some extent, the sparse matrix problem is avoided; on the other hand, the present application only needs to scan the transaction data (the first historical purchase record) once to complete the statistics of the purchase frequency of all products, compared with the traditional algorithm of generating frequent item sets, the I / O operation and the memory occupation are reduced, and thus the system resource consumption is reduced.

[0053] According to the embodiments of the present application, before the data statistics processing based on the transaction information corresponding to each of the plurality of products is performed to obtain the frequency of each of the plurality of products being purchased, the information generation method further includes: constructing a purchase list according to the first historical purchase record. Each row corresponds to a purchase transaction and contains a transaction number and a product set purchased by a user. The first historical purchase record can be data cleaned before the purchase list is constructed to remove duplicate and invalid records, and to ensure the accuracy and integrity of the data.

[0054] Exemplarily, user 1; transaction number 00023; transaction time 0512;

product A, product C…

product B, product C, product D…

product A, product C, product D…

product B, product C, product G…

product C, product G…

[0055] The data in the purchase list is converted into a vertical format. Each row corresponds to a product and transaction information of purchasing the product.

[0056] Exemplarily, product A; 【00023, 00025…】; product B; 【00024, 00026…】; product C; 【00023, 00024, 00025, 00026, 00027…】; product D; 【00024, 00025…】; product G; 【00026, 00027…】; product…

[0057] According to the embodiments of the present application, the transaction information includes a transaction number, for example, transaction number 00024, transaction number 00025. The data statistics processing based on the transaction information corresponding to each of the plurality of products is performed to obtain the frequency of each of the plurality of products being purchased, including: calculating the frequency of each product being purchased in the plurality of transaction numbers to obtain the frequency of each of the plurality of products being purchased.

[0058] For example, product A is purchased 2 times, product B is purchased 2 times, product C is purchased 5 times, product D is purchased 2 times, and product G is purchased 2 times.

[0059] Figure 3 A flowchart illustrating a method for determining a target product portfolio according to an embodiment of this application is shown.

[0060] like Figure 3 As shown, the method 300 for determining the target product portfolio includes operations S310 to S330.

[0061] In operation S310, multiple initial products are determined based on the purchase frequency of each product and the first frequency threshold.

[0062] According to embodiments of this application, products exceeding a first frequency threshold (e.g., set to 100 times) are filtered out based on their purchase frequency, forming a preliminary product set. Frequency filtering eliminates low-frequency purchases and focuses on products generally preferred by most users. For example, if the first frequency threshold is 30, and product A is purchased 50 times, then product A is identified as a preliminary product; if product B is purchased only 29 times, then product B is removed and no longer participates in the product portfolio.

[0063] When operating the S320, multiple candidate product combinations are obtained based on multiple initial product selections.

[0064] According to an embodiment of this application, in the initial selection of products, product combinations purchased simultaneously under the same transaction number are identified. That is, transaction records are analyzed; if a user purchases both product A and product C in a transaction, a candidate combination [A, C] is formed, and so on, generating all possible co-purchase combinations. For example, if a user purchases product A ("Money Market Fund") and product C ("Fixed-Term Investment"), then [A, C] becomes a candidate combination. Similarly, [A, D], [A, G], [C, D], [C, G], etc., can also be generated, but not all are listed here. However, since product B is determined not to be an initial selection product in the above example, it is deleted and no longer participates in product combinations. This is because if a product is purchased less frequently, then the frequency of that product being purchased simultaneously with other products will necessarily be less frequent.

[0065] In operation S330, at least one target product combination is determined based on multiple candidate product combinations and a second frequency threshold.

[0066] According to the embodiments of the present application, the co-occurrence frequency of the candidate product combination (i.e., how many transactions simultaneously purchase the products in the combination) is counted, and the combination exceeding the second frequency threshold is determined as the target product combination. This step filters out the high-frequency co-purchase combination, ensuring that the recommended products have strong relevance. For example, the candidate combination

A, C

C, D

[0067] According to the embodiments of the present application, the first frequency threshold and the second frequency threshold can be obtained by predicting the historical user data set through the trained target model.

[0068] According to the embodiments of the present application, the first frequency threshold filters out low-frequency products, ensuring that the preliminary selected products are generally purchased by users; the transaction number is analyzed to analyze the co-occurrence relationship of the products, and only the combination purchased simultaneously in the same transaction is retained, ensuring that the candidate combination has real business relevance. The second frequency threshold further filters the low-frequency co-purchase combination, and the target product combination has statistical significance, reducing the noise in the recommendation result. Therefore, after the first frequency threshold filtering, the number of preliminary selected products is much less than the total number of products, and the calculation amount is greatly reduced when generating the candidate combination subsequently. Compared with the traditional association rule algorithm that may generate a large number of low-frequency candidate combinations, the present application directly deletes sparse items and sparse combinations, reducing the interference of data sparsity on the recommendation effect. The second threshold directly excludes low-frequency co-purchase combinations, without the need for complex support and confidence calculations for all possible combinations, reducing memory usage and calculation time.

[0069] Figure 4 A method flowchart for determining a candidate product combination according to an embodiment of the present application is schematically shown.

[0070] As shown in Figure 4 The method 400 for obtaining multiple candidate product combinations includes operations S410-S420.

[0071] In operation S410, multiple preliminary selected products are combined two by two to obtain multiple preliminary selected product combinations.

[0072] According to the embodiments of the present application, the multiple preliminary selected products filtered out are paired two by two to form the most basic preliminary selected product combination. For example, three preliminary selected products A, C, and D generate

A, C

A, D

C, D

[0073] In operation S420, the preliminary selected product combinations satisfying the second frequency threshold are combined two by two, and so on until no combination satisfying the second frequency threshold appears, to obtain multiple candidate product combinations.

[0074] According to the embodiment of the present application, for the primary product combinations satisfying the second frequency threshold, they are further combined two by two to form higher-order combinations. For example, if

A, C

A, D

A, C, D

A, C

A, D

A, C, D

[0075] According to the embodiment of the present application, starting from the basic two-by-two combinations (such as

A, C

A, C, D

[0076] According to the embodiment of the present application, if the number of the obtained multiple candidate product combinations is less than a predetermined number, the minimum scale can be reduced when obtaining the predicted first frequency threshold by using the target model.

[0077] Figure 5 A method flowchart for determining product recommendation information according to an embodiment of the present application is schematically shown.

[0078] As shown in Figure 5 The method 500 for generating product recommendation information includes operations S510-S540.

[0079] Operation S510, determining a plurality of products that a target user has purchased according to second historical purchase records.

[0080] According to the embodiment of the present application, by analyzing the second historical purchase records of the target user, all the products that the user has purchased are sorted out. For example, from the second transaction records, it is extracted that the target user A has purchased two products of "money fund" and "fixed-term financing".

[0081] Operation S520, determining at least one reference product combination from the target product combinations according to the plurality of products that the target user has purchased.

[0082] According to an embodiment of the present application, in the determined target product combination, a combination containing products purchased by the target user is screened out as a reference. For example, the target user A has purchased the "insurance" product, and the target product combination containing the "insurance" product is determined as the reference product combination, for example, the reference product combination 1 [money fund, fixed-term financing, insurance].

[0083] Operation S530, determining the products not purchased by the target user according to the product information of the reference product combination.

[0084] According to an embodiment of the present application, the products purchased by the user are excluded from the reference product combination, and the remaining products are the recommended candidates. For example, in the reference combination [money fund, fixed-term financing, insurance], the user has purchased insurance, and the money fund and fixed-term financing are the products not purchased by the target user.

[0085] Operation S540, generating product recommendation information for the target user according to the products not purchased by the target user.

[0086] According to an embodiment of the present application, the products not purchased in the candidate pool are sorted into a recommendation list, and the recommendation content for the user is generated in combination with the product information (such as yield, risk level, etc.). For example, the "fixed-term financing" product is recommended to the user, which can include its guarantee range and adaptation scene description, etc.

[0087] According to an embodiment of the present application, the information generation method further comprises: obtaining first attribute information of a plurality of historical users and second attribute information of a plurality of target users; classifying the plurality of historical users according to the first attribute information to obtain a plurality of historical user groups; classifying the plurality of target users according to the second attribute information to obtain a plurality of target user groups.

[0088] According to an embodiment of the present application, the first attribute information of a plurality of historical users, such as age, and the second attribute information of a plurality of target users (the attribute type is consistent with that of the historical users) are collected, and the historical users or target users are classified according to age, so as to mine the correlation of purchasing products according to different types of user groups. For example, the age range of the historical users includes 0-25 years old; 26-45 years old; 46-65 years old; 65 years old and above; if the age of the target user is 26-45 years old, the first historical purchase record corresponding to the historical user of 26-45 years old and the first frequency threshold are used to recommend products for the target user.

[0089] Further, the attribute information can also be occupation, income, risk preference, etc. Embodiments of the present application can generate recommendation information according to only one attribute information, or can generate recommendation information according to a combination of multiple attribute information. For example, the historical user attribute includes "30 years old + white-collar + annual income 200,000 + stable type", and the target user attribute includes "25 years old + student + annual income 50,000 + conservative type", etc.

[0090] According to embodiments of the present application, if the recommendation information is generated according to the combination of multiple attribute information, the users with similar attributes can also be divided into the same group based on the first attribute information of the historical user. For example, the historical users with "age 30-40, annual income 15-30, stable type" are classified into "middle-class stable type group", and the historical users with "age 20-25, annual income less than 50,000, conservative type" are classified into "young conservative type group". Among them, the clustering algorithm (such as K-means) or rule matching (such as according to age interval and income level) can be used to classify the users.

[0091] According to embodiments of the present application, the second attribute information of the target user is processed by using the same attribute dimension and classification rule, and the target users with similar attributes are classified into the corresponding historical user group category. For example, if the target user meets "25 years old + student + annual income 50,000 + conservative type", it is classified into "young conservative type group", and the mapping between the target user and the historical user group is realized.

[0092] According to embodiments of the present application, the users are classified by using multi-dimensional attribute information such as age, income, and risk preference, and a refined user group label is formed. For example, in the financial scenario, the users can be classified into "high-income aggressive type" and "medium-income balanced type" by combining income and risk preference, and different product strategies are adapted. If the business requirement is adjusted (such as adding "occupation" as a classification dimension), the attribute field can be flexibly extended and re-clustered, without the need to restructure the underlying logic, and the scalability of user grouping is improved.

[0093] According to embodiments of the present application, the product recommendation information for the multiple target users is generated according to the second historical purchase record of the multiple target users and the product information of the target product combination, including: in the case that at least one target user group is the same as at least one historical user group, generating product recommendation information for at least one target user according to the second historical purchase record of at least one target user and the product information of the target product combination corresponding to at least one historical user group.

[0094] According to an embodiment of the present application, when the attribute characteristics of the target user group and the historical user group are completely or highly coincident, it is determined that the groups are the same. For example, the target user A belongs to the "young middle-class stable type" group, which already exists in the historical users. The second historical purchase record of the target user is obtained (for example, the user A has purchased "money fund" and "fixed-term financing", the target product combination corresponding to the historical group (such as "money fund + fixed-term financing + index fund") is called. Then the "index fund" product is directly recommended to the target user A.

[0095] According to an embodiment of the present application, by integrating the high-frequency product combination of the historical group and the individual purchase data of the target user, the recommendation accuracy and efficiency are improved. On the one hand, the un-purchased items in the combination of the purchased products of the user are combined to avoid blind pushing of the full set, and the recommendation accuracy is improved. For example, if the group combination contains 3 products, and the user has purchased 2, the accuracy of recommending the remaining 1 product is higher than directly pushing the full set of the group combination. Without the need to calculate the product correlation degree for each target user, the high-frequency combination of the historical group is directly reused, and the recommendation efficiency and accuracy are improved. On the other hand, for new users, if their attributes match the historical group, the group combination can be directly recommended, which can to some extent avoid the cold start problem, make up for the recommendation deviation caused by sparse individual data, and realize stable recommendation in the data sparse scene. For example, the new user B belongs to the "retirement stable type" group, and the high-frequency product combination of the group is directly recommended to solve the problem of no data reference in the cold start stage. When the purchase record of the target user is less (such as only purchasing 1 product), the group combination can be used as a supplementary feature to avoid the deviation of the recommendation result caused by insufficient individual data.

[0096] According to an embodiment of the present application, in the process of training the target model according to the historical data, the model can also be trained according to different user groups. Specifically, the historical sample data and labels of the conservative user group are used to train the first model to obtain the first target model. After inputting the first historical purchase record of the historical user belonging to the conservative user group into the first target model, the first frequency threshold corresponding to the conservative user group is output, for example, 10%. The historical sample data and labels of the aggressive user group are used to train the second model to obtain the second target model. After inputting the first historical purchase record of the historical user belonging to the aggressive user group into the first target model, the first frequency threshold corresponding to the conservative user group is output, for example, 5%. This shows that compared with the conservative user group, the aggressive user group is more willing to try other products.

[0097] Based on the above information generation method, the present application also provides an information generation device. The following will be combined with the description of the information generation method to describe the information generation device. Figure 6 The device will be described in detail.

[0098] Figure 6A structural block diagram of an information generation apparatus according to an embodiment of the present application is shown schematically.

[0099] As shown in Figure 6 The information generation apparatus 600 of this embodiment includes a first acquisition module 610, a statistics module 620, a determination module 630, and a generation module 640.

[0100] The first acquisition module 610 is configured to acquire a plurality of first historical purchase records of a plurality of products purchased by a plurality of historical users, wherein the first historical purchase records include transaction information corresponding to the plurality of products respectively. In an embodiment, the first acquisition module 610 can be configured to perform the operation S210 described above, and details are not repeated here.

[0101] The statistics module 620 is configured to perform data statistics processing based on the transaction information corresponding to the plurality of products respectively, to obtain a frequency of purchase of the plurality of products respectively. In an embodiment, the statistics module 620 can be configured to perform the operation S220 described above, and details are not repeated here.

[0102] The determination module 630 is configured to determine a target product combination according to the frequency of purchase of the plurality of products respectively. In an embodiment, the determination module 630 can be configured to perform the operation S230 described above, and details are not repeated here.

[0103] The generation module 640 is configured to generate product recommendation information for a plurality of target users according to second historical purchase records of the plurality of target users and product information of the target product combination. In an embodiment, the generation module 640 can be configured to perform the operation S240 described above, and details are not repeated here.

[0104] According to an embodiment of the present application, the target product combination is obtained according to the purchase frequency of the historical users, i.e. the correlation between products and products is obtained, so that even if the historical data of the target user is limited, e.g. the target user has only purchased one product, the target user can still be recommended according to the historical user preferences, and to some extent, the problem of sparse matrix is avoided. On the other hand, the present application only needs to scan the transaction data (first historical purchase records) once to complete the statistics of the purchase frequency of all products, compared with the traditional algorithm of generating frequent item sets which needs to scan repeatedly, the I / O operation and memory occupation are reduced, and thus the system resource consumption is reduced.

[0105] According to an embodiment of the present application, the information generation apparatus further includes a second acquisition module, a first classification module, and a second classification module.

[0106] The second obtaining module is configured to obtain first attribute information of a plurality of historical users and second attribute information of a plurality of target users; the first classification module is configured to classify the plurality of historical users according to the first attribute information to obtain a plurality of historical user groups; and the second classification module is configured to classify the plurality of target users according to the second attribute information to obtain a plurality of target user groups.

[0107] According to an embodiment of the present application, the generating module 640 comprises a generating sub-module.

[0108] The generating sub-module is configured to, in a case where at least one target user group is identical to at least one historical user group, generate product recommendation information for at least one target user according to a second historical purchase record of the at least one target user and product information of a target product combination corresponding to the at least one historical user group.

[0109] According to an embodiment of the present application, the transaction information comprises a transaction number, and the statistical module 620 comprises a calculating sub-module.

[0110] The calculating sub-module is configured to calculate a frequency at which each product is purchased in the plurality of transaction numbers to obtain a frequency at which each product is purchased.

[0111] According to an embodiment of the present application, the determining module 630 comprises a first determining sub-module, an obtaining sub-module, and a second determining sub-module.

[0112] The first determining sub-module is configured to determine a plurality of preliminary selected products according to the frequency at which each product is purchased and a first frequency threshold; the obtaining sub-module is configured to obtain a plurality of candidate product combinations according to the plurality of preliminary selected products; and the second determining sub-module is configured to determine at least one target product combination according to the plurality of candidate product combinations and a second frequency threshold.

[0113] According to an embodiment of the present application, the obtaining sub-module comprises a first combination unit and a second combination unit.

[0114] The first combination unit is configured to combine the plurality of preliminary selected products two by two to obtain a plurality of preliminary selected product combinations; and the second combination unit is configured to combine preliminary selected product combinations that satisfy the second frequency threshold two by two, and to continue this process until no combination that satisfies the second frequency threshold appears, to obtain the plurality of candidate product combinations.

[0115] According to an embodiment of the present application, the generating module 640 comprises a third determining sub-module, a fourth determining sub-module, a fifth determining sub-module, and a generating sub-module.

[0116] The third determining sub-module is configured to determine a plurality of products that the target user has purchased according to the second historical purchase record; the fourth determining sub-module is configured to determine at least one reference product combination from the target product combination according to the plurality of products that the target user has purchased; the fifth determining sub-module is configured to determine a product that the target user has not purchased according to product information of the reference product combination; and the generating sub-module is configured to generate product recommendation information for the target user according to the product that the target user has not purchased.

[0117] According to embodiments of the present application, any of the first obtaining module 610, the statistical module 620, the determining module 630 and the generating module 640 can be combined in one module, or any of them can be split into multiple modules. Alternatively, at least part of the function of one or more of these modules can be combined with at least part of the function of other modules, and implemented in one module. According to embodiments of the present application, at least one of the first obtaining module 610, the statistical module 620, the determining module 630 and the generating module 640 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 chip, a system on board, a system on package, an application specific integrated circuit (ASIC), or any other reasonable way of hardware or firmware that can be integrated or packaged, or implemented in any one of software, hardware and firmware or in a proper combination of any of them. Alternatively, at least one of the first obtaining module 610, the statistical module 620, the determining module 630 and the generating module 640 can be at least partially implemented as a computer program module which can perform corresponding functions when it is run.

[0118] Figure 7 A block diagram of an electronic device suitable for implementing the information generation method according to embodiments of the present application is schematically shown.

[0119] As Figure 7 shown, the electronic device 700 according to embodiments of the present application includes a processor 701 which can perform various appropriate actions and processes according to programs stored in a read only memory (ROM) 702 or loaded from a storage portion 708 into a random access memory (RAM) 703. The processor 701 can 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 (such as an application specific integrated circuit (ASIC)), etc. The processor 701 can also include an on-board memory for cache use. The processor 701 can include a single processing unit or multiple processing units for performing different actions of the method processes according to embodiments of the present application.

[0120] In the RAM 703, various programs and data required for the operation of the electronic device 700 are stored. The processor 701, the ROM 702, and the RAM 703 are connected to each other via the bus 704. The processor 701 performs various operations of the method flow according to the embodiments of the present application by executing the programs in the ROM 702 and / or the RAM 703. It should be noted that the programs can also be stored in one or more memories other than the ROM 702 and the RAM 703. The processor 701 can also perform various operations of the method flow according to the embodiments of the present application by executing the programs stored in the one or more memories.

[0121] According to the embodiments of the present application, the electronic device 700 can further include an input / output (I / O) interface 705, which is also connected to the bus 704. The electronic device 700 can further include one or more of the following components connected to the input / output (I / O) interface 705: an input part 706 including a keyboard, a mouse, and the like; an output part 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), and the like, and a speaker, and the like; a storage part 708 including a hard disk, and the like; and a communication part 709 including a network interface card such as a LAN card, a modem, and the like. The communication part 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the input / output (I / O) interface 705 as necessary. A removable medium 711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, and the like is mounted on the drive 710 as necessary, so that a computer program read therefrom is installed in the storage part 708 as necessary.

[0122] The present application also provides a computer readable storage medium, which can be included in the device / apparatus / system described in the above embodiments; or can exist separately without being assembled into the device / apparatus / system. The above computer readable storage medium carries one or more programs, when the one or more programs are executed, the method according to the embodiments of the present application is implemented.

[0123] According to an embodiment of the present application, the computer readable storage medium can be a non-transitory computer readable storage medium, for example, can include but not limited to: a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In this application, a computer readable storage medium can be any tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. For example, according to an embodiment of the present application, the computer readable storage medium can include the ROM 702 and / or the RAM 703 described above and / or one or more memory other than the ROM 702 and the RAM 703.

[0124] Embodiments of the present application also include a computer program product, which includes a computer program containing program codes for executing the methods shown in the flowcharts. When the computer program product is run in a computer system, the program codes are used to make the computer system implement the information generation method provided by the embodiments of the present application.

[0125] The above functions defined in the system / device / apparatus of the embodiments of the present application are performed when the computer program is executed by the processor 701. According to an embodiment of the present application, the system, device, module, unit, etc. described above can be implemented by computer program modules.

[0126] In one embodiment, the computer program can rely on tangible storage media such as optical storage media, magnetic storage media, etc. In another embodiment, the computer program can also be transmitted, distributed, and downloaded in the form of signals on a network medium, and be downloaded and installed through the communication part 709, and / or installed from the detachable medium 711. The program codes contained in the computer program can be transmitted by any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.

[0127] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 709, and / or installed from the detachable medium 711. When the computer program is executed by the processor 701, the above functions defined in the system of the embodiments of the present application are performed. According to an embodiment of the present application, the system, device, apparatus, module, unit, etc. described above can be implemented by computer program modules.

[0128] According to embodiments of the present application, program code for implementing the computer programs provided by embodiments of the present application can be written in any combination of one or more programming languages, and can be implemented in a high-level procedural and / or object-oriented programming language, and / or in assembly / machine language. Programming languages include, but are not limited to, Java, C++, python, "C", or the like. Program code can execute entirely on a user's computing device, partly on the user's device, as a stand-alone software package, partly on a remote computing device, or entirely on the remote computing device or server. In the latter scenario, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computing device, such as through the Internet using an Internet Service Provider.

[0129] The computer program instructions can also be loaded onto a computer or other programmable information processing apparatus to cause a series of operations to be performed on the computer or other programmable information processing apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable information processing apparatus implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0130] Those skilled in the art will appreciate that features recited in the various embodiments of the present application can be combined and / or integrated in various combinations, even if such combinations have not been explicitly recited in the present application. In particular, the features recited in the various embodiments of the present application can be combined and / or integrated in various combinations, without departing from the spirit and teachings of the present application. All such combinations are within the scope of the present application.

Claims

1. An information generation method characterized by comprising: The method comprises: obtaining a plurality of first historical purchase records of a plurality of products purchased by a plurality of historical users, wherein the first historical purchase records comprise transaction information corresponding to the plurality of products respectively; based on the transaction information corresponding to the plurality of products respectively, performing data statistical processing to obtain the frequency of purchase of the plurality of products respectively; determining a target product combination according to the frequency of purchase of the plurality of products respectively; generating product recommendation information for a plurality of target users according to second historical purchase records of the plurality of target users and product information of the target product combination.

2. The method of claim 1, wherein, The method further comprises: obtaining first attribute information of the plurality of historical users and second attribute information of the plurality of target users; classifying the plurality of historical users according to the first attribute information to obtain a plurality of historical user groups; classifying the plurality of target users according to the second attribute information to obtain a plurality of target user groups.

3. The method of claim 2, wherein, Generating product recommendation information for a plurality of target users according to second historical purchase records of the plurality of target users and product information of the target product combination comprises: in the case that at least one target user group is the same as at least one historical user group, generating product recommendation information for the at least one target user according to the second historical purchase records of the at least one target user and the product information of the target product combination corresponding to the at least one historical user group.

4. The method of claim 1, wherein, The transaction information comprises a transaction number, Based on the transaction information corresponding to the plurality of products respectively, performing data statistical processing to obtain the frequency of purchase of the plurality of products respectively comprises: calculating the frequency of purchase of each product in a plurality of transaction numbers to obtain the frequency of purchase of the plurality of products respectively.

5. The method of claim 1, wherein, Determining a target product combination according to the frequency of purchase of the plurality of products respectively comprises: determining a plurality of preliminary products according to the frequency of purchase of the plurality of products respectively and a first frequency threshold; obtaining a plurality of candidate product combinations according to the plurality of preliminary products; determining the at least one target product combination according to the plurality of candidate product combinations and a second frequency threshold.

6. The method of claim 5, wherein, Obtaining a plurality of candidate product combinations according to the plurality of preliminary products comprises: combining the plurality of preliminary products two by two to obtain a plurality of preliminary product combinations; combining preliminary product combinations that satisfy the second frequency threshold two by two, and so on until no combination that satisfies the second frequency threshold appears, to obtain the plurality of candidate product combinations.

7. The method of claim 1, wherein, Generating product recommendation information for a plurality of target users according to second historical purchase records of the plurality of target users and product information of the target product combination comprises: determining a plurality of products purchased by the target user according to the second historical purchase records; determining at least one reference product combination from the target product combination according to the plurality of products purchased by the target user; determining products not purchased by the target user according to product information of the reference product combination; generating product recommendation information for the target user according to the products not purchased by the target user.

8. An information generation apparatus characterized by comprising: The device comprises: The first obtaining module is configured to obtain a first historical purchase record of a plurality of products purchased by a plurality of historical users, wherein the first historical purchase record comprises transaction information corresponding to each of the plurality of products; The statistical module is configured to perform data statistical processing based on the transaction information corresponding to each of the plurality of products, to obtain a frequency of purchase of each of the plurality of products; The determining module is configured to determine a target product combination according to the frequency of purchase of each of the plurality of products; The generating module is configured to generate product recommendation information for a plurality of target users according to second historical purchase records of the plurality of target users and product information of the target product combination.

9. An electronic device comprising: one or more processors; memory for storing one or more computer programs, 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 1-7.

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

11. A computer program product comprising computer programs or instructions, characterized in that, The computer program or instructions are executed by the processor to implement the steps of the method according to any one of claims 1-7. The computer program or instructions are executed by the processor to implement the steps of the method according to any one of claims 1-7.

Images