Commodity recommendation method and device, electronic equipment and readable storage medium
By obtaining and analyzing user's shopping data, calculating product recommendation index and filtering levels using the recommendation index algorithm, the problem of low accuracy of product recommendation in the existing technology is solved, personalized and highly accurate product recommendations are achieved, and user experience is improved.
Patent Information
- Application Number
- CN202410009672.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-02
- Publication Date
- 2025-07-04
AI Technical Summary
The existing product recommendation methods are based on keyword matching and cannot effectively identify user intentions, resulting in low recommendation accuracy and prone to deviations.
By obtaining the user's shopping data, analyzing browsing and purchasing records, using the recommendation index algorithm to calculate the product recommendation index, and filtering out product information that meets user needs based on the level threshold, and entering the product matching model for recommendation.
It improves the accuracy of product recommendations, reduces recommendation bias, improves users' shopping experience, and provides more personalized and product recommendations that meet user needs.
Smart Images
Figure CN120258923A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of information recommendation, and particularly to a product recommendation method, device, electronic device, and readable storage medium. Background Art
[0002] With the rapid rise of major e-commerce platforms, online shopping has become the most mainstream shopping method currently. Usually, a large amount of search and query of the massive product data displayed by merchants is required for online shopping to lock in the purchase target. Therefore, as shopping platforms gradually mature, users have higher and higher requirements for product search recommendations. Currently, in the online shopping scenario, intelligent recommendation is still in its infancy. Recommending similar products to users based on their shopping behavior or search behavior can meet users' shopping needs and provide products that better suit their needs, which can promote users' shopping interest.
[0003] Existing product search recommendation methods mainly rely on graphs or use methods such as text keyword matching. It is necessary to analyze the query conditions input by users to obtain keywords, and then perform keyword matching between the keywords and all fields in the product graph or all fields in the indexed text. After the search results hit the products corresponding to the fields, the product information results are displayed according to the matching scores and the reordering conditions set manually.
[0004] However, the above keyword matching-based method, due to focusing more on keyword matching, cannot well identify user intentions, rarely filters miscellaneous data and analyzes user behavior, so it is prone to recommendation deviation, resulting in low accuracy of product recommendation. Summary of the Invention
[0005] This application provides a product recommendation method, device, electronic device, and readable storage medium to solve the technical problem that the existing product recommendation method has low recommendation accuracy and is prone to recommendation deviation.
[0006] According to the first aspect disclosed in this application, this application provides a product recommendation method, including:
[0007] Obtain the shopping data of the target user; wherein, the shopping data includes browsing records and purchase records;
[0008] Parse the shopping data to obtain multiple groups of product information; wherein, the product information includes multiple search product information and purchase product information belonging to the same product category;
[0009] Based on the recommendation index algorithm, parse each group of product information to obtain the product recommendation index of each group of product information;
[0010] Classify the recommendation index of each product to obtain the recommendation level of each group of product information;
[0011] For each group of product information, if the recommendation level of the product information is greater than a preset level threshold, the product information is input into a product matching model to obtain recommended products.
[0012] In a feasible implementation, the shopping data is parsed to obtain multiple sets of product information, including:
[0013] Based on the browsing history, a plurality of search product information is obtained; wherein the search product information includes the search product information category, the search product information price and the product browsing time;
[0014] Based on the purchase record, a plurality of purchased commodity information is obtained; wherein the purchased commodity information includes the purchased commodity information category and the purchased commodity information price;
[0015] The search product information and purchase product information belonging to the same product category are filtered and grouped to obtain multiple groups of product information.
[0016] In a feasible implementation manner, the method further includes:
[0017] For each searched product information, if the product browsing time of the searched product information is not greater than a preset time threshold, the searched product information is eliminated.
[0018] In a feasible implementation manner, the method further includes:
[0019] The recommendation index algorithm satisfies the following formula:
[0020]
[0021] Where Jzs represents the recommended index of the product, SPsjz represents the actual index of the product, SPckz represents the reference index of the product, a represents the weight coefficient of the difference between the reference index of the product and the actual index of the product, and 0 <a<1;CZm表示商品的时间差值指数,b表示时间差值指数的权重系数,且0<b<1;C表示校准指数。
[0022] In a feasible implementation manner, the method further includes:
[0023] The commodity reference index satisfies the following formula:
[0024]
[0025] Wherein, SPckz represents the commodity reference index, SSn represents the number of search commodity information in the commodity information, JG1, JG2, JG3, ..., JGn represent the price of each search commodity information in the commodity information, and D represents the first correction constant;
[0026] The actual index of the commodity satisfies the following formula:
[0027]
[0028] Wherein, SPsjz represents the actual index of the commodity, GMn represents the number of purchased commodity information in the commodity information, GMJG1, GMJG2, GMJG3, ..., GMJGn represent the price of each purchased commodity information in the commodity information, and E represents the second correction constant;
[0029] The time difference index satisfies the following formula:
[0030]
[0031] Wherein, CZm represents the time difference index, F represents the third correction constant, TLm represents the browsing time of searching for product information, r represents the decay value, and t represents the time interval from the start of searching for product information to the appearance of search results.
[0032] In a feasible implementation, each commodity recommendation index is graded to obtain the recommendation level of each group of commodity information, including:
[0033] For each product recommendation index, compare the product recommendation index with a preset first threshold and a second threshold; wherein the first threshold is greater than the second threshold;
[0034] If the product recommendation index is greater than the first threshold, determining that the product recommendation index is a first recommendation level;
[0035] If the product recommendation index is less than the first threshold and the product recommendation index is greater than the second threshold, determining that the product recommendation index is a second recommendation level; wherein the second recommendation index is above the first recommendation level;
[0036] If the product recommendation index is less than the second threshold, the product recommendation index is determined to be a third recommendation level; wherein the third recommendation level is less than the second recommendation level.
[0037] In a feasible implementation manner, the method further includes:
[0038] If the recommendation level of the product information is lower than the preset level threshold, the product information will be discarded.
[0039] According to a second aspect disclosed in the present application, the present application provides a commodity recommendation device, including:
[0040] A data acquisition module for acquiring the shopping data of a target user; wherein, the shopping data includes browsing records and purchase records;
[0041] A data parsing module for parsing the shopping data to obtain multiple groups of commodity information; wherein, the commodity information includes multiple search commodity information and purchase commodity information belonging to the same commodity category;
[0042] A recommendation index module for parsing each group of commodity information based on a recommendation index algorithm to obtain the commodity recommendation index of each group of commodity information;
[0043] An index grading module for grading each commodity recommendation index to obtain the recommendation level of each group of commodity information;
[0044] A commodity matching module for, for each group of commodity information, if the recommendation level of the commodity information is greater than a preset level threshold, inputting the commodity information into a commodity matching model to obtain recommended commodities.
[0045] According to a third aspect disclosed in the present application, there is provided an electronic device, including a processor and a memory communicatively connected to the processor;
[0046] The memory stores computer execution instructions;
[0047] The processor executes the computer execution instructions stored in the memory to implement the method according to any one of the first aspects.
[0048] According to a fourth aspect disclosed in the present application, there is provided a computer-readable storage medium storing computer execution instructions, and when the computer execution instructions are executed by a processor, they are used to implement the method according to any one of the first aspects.
[0049] According to a fifth aspect disclosed in the present application, there is provided a computer program product including a computer program, and when the computer program is executed by a processor, it is used to implement the method according to any one of the first aspects.
[0050] Compared with the prior art, the present application has the following beneficial effects:
[0051] A product recommendation method, device, electronic device, and readable storage medium provided by this application obtain the shopping data of a target user; wherein, the shopping data includes browsing records and purchase records; parse the shopping data to obtain multiple groups of product information; wherein, the product information includes multiple search product information and purchase product information belonging to the same product category; based on a recommendation index algorithm, parse each group of product information to obtain the product recommendation index of each group of product information; classify each product recommendation index to obtain the recommendation level of each group of product information; for each group of product information, if the recommendation level of the product information is greater than a preset level threshold, input the product information into a product matching model to obtain the recommended products. This technical means provides more personalized and user-demand-compliant product recommendations, improves the accuracy of product recommendations, reduces the deviation of product recommendations, and enhances the user's shopping experience. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with this application, and are used together with the specification to explain the principles of this application.
[0053] Figure 1 It is a schematic flowchart of a product recommendation method provided by an embodiment of this application;
[0054] Figure 2 It is a schematic flowchart of another product recommendation method provided by an embodiment of this application;
[0055] Figure 3 It is a schematic structural diagram of a product recommendation device provided by an embodiment of this application;
[0056] Figure 4 It is a schematic structural diagram of an electronic device provided by an embodiment of this application.
[0057] Through the above accompanying drawings, specific embodiments of this application have been shown, and there will be more detailed descriptions hereinafter. These drawings and textual descriptions are not intended to limit the scope of the concept of this application in any way, but to illustrate the concept of this application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0058] Here, exemplary embodiments will be described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Instead, they are merely examples of devices and methods consistent with some aspects of this application as detailed in the appended claims.
[0059] With the rapid rise of major e-commerce platforms, online shopping has become the most mainstream shopping method. Usually, a large amount of searching and querying of the massive product data displayed by merchants is required for online shopping to lock in the purchase target. Therefore, as shopping platforms gradually mature, users have higher and higher requirements for product search recommendations. Currently, in the online shopping scenario, intelligent recommendations are still in their infancy. Recommending similar products to users based on their shopping behavior or search behavior to meet users' shopping needs and provide products that better suit their needs can boost users' shopping interest.
[0060] Existing product search recommendation methods mainly rely on knowledge graphs or use methods such as text keyword matching. They need to analyze the query conditions entered by users to obtain keywords, and then perform keyword matching between the keywords and all fields in the product knowledge graph or all fields in the indexed text. After the search results hit the products corresponding to the fields, the product information results are displayed according to the matching scores and the reordering conditions set manually.
[0061] However, the above keyword matching-based methods, due to their greater emphasis on keyword matching, cannot well identify users' intentions, rarely screen miscellaneous data and analyze users' behaviors. Therefore, it is easy to have recommendation deviations, resulting in low accuracy of product recommendations.
[0062] To address the above technical problems, this application proposes a product recommendation method. By collecting and analyzing users' shopping data, it provides more personalized product recommendations that meet users' needs, improves the accuracy of product recommendations, reduces product recommendation deviations, and enhances users' shopping experience.
[0063] The technical solution of the product recommendation method provided by this application will be described in detail below through specific embodiments. It should be noted that the following embodiments can exist independently or be combined with each other. For the same or similar content, it may not be repeated in different embodiments.
[0064] Figure 1 is a flowchart of a product recommendation method provided by an embodiment of this application. Refer to Figure 1 , in some embodiments, the process of the product recommendation method includes the following steps:
[0065] S101, obtain the shopping data of the target user; wherein, the shopping data includes browsing records and purchase records.
[0066] Among them, the browsing records and purchase records of users contain the products searched, browsed, and purchased by users, which can reflect which products users are interested in. Therefore, more targeted product recommendations can be made to users through this shopping data.
[0067] S102, Parse the shopping data to obtain multiple groups of product information; among them, the product information includes multiple search product information and purchase product information belonging to the same product category; among them, the product information includes multiple search product information and purchase product information belonging to the same product category.
[0068] Among them, when parsing the shopping data, there are multiple search product information and purchase product information in the shopping data, which are grouped by product category, and the search product information and purchase product information belonging to the same category are grouped into the same product information.
[0069] S103, Parse each group of product information based on the recommendation index algorithm to obtain the product recommendation index of each group of product information.
[0070] Among them, the product recommendation index of the product information is obtained through the recommendation index algorithm, and the product recommendation index can reflect the recommendation value of the products in this category.
[0071] S104, Grade each product recommendation index to obtain the recommendation level of each group of product information.
[0072] Among them, the corresponding recommendation level is obtained based on the product recommendation index, and the purpose is to screen the product information according to different recommendation levels.
[0073] Among them, through the grading of the product recommendation index, the recommendation value of the product is evaluated and graded. Then, based on these grading comparison results, it is judged whether the product information is worthy of being used as the input data of the product matching model, so as to exclude recommendations with low recommendation value or those that do not meet the requirements.
[0074] S105, For each group of product information, if the recommendation level of the product information is greater than the preset level threshold, input the product information into the product matching model to obtain recommended products.
[0075] Among them, a product matching model is pre-set to obtain recommended products that are associated and matched with the product information through the product matching model.
[0076] Specifically, the product matching model can be a deep matching model or other recommendation algorithms. The product matching model can analyze the user's interests, preferences and purchase behaviors based on the product information, and recommend relevant products according to this information, thereby improving the user experience and sales conversion rate of the shopping application APP or shopping website. This personalized product recommendation method can increase the user's stickiness to the shopping application APP or shopping website, thus bringing more commercial value. In actual applications, the product matching model can, on the basis of data processing, combine deep learning technology to learn the complex relationship between products and users from massive data to achieve more intelligent product search and recommendation.
[0077] Specifically, the deep matching model is a deep learning-based technology used to solve matching problems in natural language processing. It is mainly used to judge the similarity or matching degree between two texts, such as the matching between questions and answers, and the relevance between queries and documents in search engines. The deep matching model is usually constructed based on neural networks, including convolutional neural networks (CNNs) and recurrent neural networks (RNNs), etc. These models extract a series of features from the input text and map them into a fixed-dimensional representation vector space, and then use these vectors to calculate the similarity or judge the matching between texts. Common deep matching models include Siamese networks, MatchZoo, DSSM, etc. By learning the semantic and context information of texts, these models can capture the semantic similarity between texts to a certain extent, thus improving the accuracy of matching. The deep matching model has a wide range of applications in natural language processing tasks, such as question answering systems, text matching, recommendation systems, etc. They can help machines understand and process human language, and improve the effect of text understanding and matching.
[0078] Specifically, after obtaining the recommended products, a product recommendation list can be generated based on the recommended products and sent to the display page of the shopping application APP or shopping website for users to browse.
[0079] In this embodiment, from the collection of user behavior data to the generation of the final product recommendation, through refined data processing, calculation, and comparison, more personalized and user-demand-compliant product recommendations are provided, improving the customer shopping experience.
[0080] On the basis of Figure 1 the embodiments shown, the technical solutions of the above product recommendation method will be further introduced below in combination with Figure 2 ...
[0081] Figure 2 FIG. Figure 2 is a schematic flowchart of another product recommendation method provided by an embodiment of the present application. Referring to
[0082] S201, obtain the shopping data of the target user; wherein, the shopping data includes browsing records and purchase records.
[0083] S202, based on the browsing records, obtain multiple search product information; wherein, the search product information includes search product information categories, search product information prices, and product browsing times.
[0084] Among them, from the user's browsing records, relevant information about the products searched and browsed by the user can be extracted, and the information of the search product information includes the category, price, and browsing time of the search product information.
[0085] Specifically, the product browsing time can be obtained from the user information in the shopping APP.
[0086] S203. For each piece of searched product information, if the product browsing time of the searched product information is not greater than the preset duration threshold, then the searched product information is excluded.
[0087] Among them, if the user's browsing time for a product is too short, it means that the user is not interested in this product. Therefore, by setting a duration threshold to screen the searched product information, only the information of the searched product information with a browsing time greater than the duration threshold can be adopted. This can avoid interference from incorrect data in the recommendation results, enabling the product recommendation model to make more accurate product recommendations based on more accurate product information.
[0088] For example, when the user browses a certain product, if the browsing time of this product is 25 seconds and the preset screening duration threshold in the shopping APP system background is 15 seconds, at this time the browsing time is greater than the duration threshold, so the system determines it as a correct data source and the system proceeds to the next step. If the browsing time of this product is 5 seconds, at this time the browsing time is less than the duration threshold, the system will determine it as an incorrect data source and then exclude this data.
[0089] In addition, other screening conditions can also be set to screen and exclude incorrect data in the searched product information and purchased product information.
[0090] S204. Based on the purchase records, obtain multiple pieces of purchased product information; among them, the purchased product information includes the category of the purchased product information and the price of the purchased product information.
[0091] Among them, from the user's purchase records, relevant information about the products purchased by the user can be extracted. The information of the purchased product information includes the category and price of the purchased product information.
[0092] S205. Screen and group the searched product information and purchased product information belonging to the same product category to obtain multiple groups of product information.
[0093] Among them, for the above-mentioned searched product information and purchased product information, products of the same product category are classified into the same group to obtain multiple groups of product information; among them, the product information includes multiple pieces of searched product information and purchased product information belonging to the same product category.
[0094] S206. Analyze each group of product information based on the recommendation index algorithm to obtain the product recommendation index of each group of product information.
[0095] Preferably, the recommendation index algorithm satisfies the following formula:
[0096]
[0097] Among them, Jzs represents the product recommendation index, SPsjz represents the actual product index, SPckz represents the reference product index, a represents the weight coefficient of the difference between the reference product index and the actual product index, and 0 < a < 1; CZm represents the time difference index of the product, b represents the weight coefficient of the time difference index, and 0 < b < 1; C represents the calibration index.
[0098] Among them, the product recommendation index and the actual product index are calculated based on the quantity and price of products searched and purchased by users for a certain type of product. These indexes can be used for product recommendation to provide more recommendation results that meet user needs and purchase behaviors. The time difference index corresponds to the sensitivity of users to time when searching for product information or content. By calculating the weight distribution of the above three parameters respectively, the above formula can reflect the recommendability of products to a certain extent under the condition that users adjust the weight coefficient and the calibration index.
[0099] Specifically, a and b are used to adjust the influence of the difference between the reference product index and the actual product index on the product recommendation index, and C is used to overall calibrate the product recommendation index.
[0100] Specifically, the specific values of a, b, and C are pre-adjusted and set according to the actual situation.
[0101] Specifically, the actual product index satisfies the following formula:
[0102]
[0103] Among them, SPsjz represents the actual product index, GMn represents the quantity of purchased product information in the product information, GMJG1, GMJG2, GMJG3,..., GMJGn represent the prices of each purchased product information in the product information, and E represents the second correction constant.
[0104] Specifically, the actual product index satisfies the following formula:
[0105]
[0106] Among them, SPsjz represents the actual product index, GMn represents the quantity of purchased product information in the product information, GMJG1, GMJG2, GMJG3,..., GMJGn represent the prices of each purchased product information in the product information, and E represents the second correction constant.
[0107] Specifically, the time difference index satisfies the following formula:
[0108]
[0109] Among them, CZm represents the time difference index, F represents the third correction constant, TLm represents the browsing time for searching product information, r represents the decay value, and t represents the time interval from the start of searching product information to the appearance of search results.
[0110] Among them, the time difference index CZm is used to consider the user's sensitivity to time when searching for product information or content. As the time interval increases, the time difference index will gradually decrease according to the setting of the decay value r. The correction constant F and the decay value r are adjusted and set in advance to flexibly optimize and adapt the time difference index according to business requirements and user behavior, ensuring that the recommended results are more in line with the user's interests and real-time needs.
[0111] S207. For each product recommendation index, compare the product recommendation index with a preset first threshold and a second threshold; among them, the first threshold is greater than the second threshold.
[0112] S208. If the product recommendation index is greater than the first threshold, determine that the product recommendation index is the first recommendation level.
[0113] S209. If the product recommendation index is less than the first threshold and the product recommendation index is greater than the second threshold, determine that the product recommendation index is the second recommendation level; among them, the second recommendation index is lower than the first recommendation level.
[0114] S210. If the product recommendation index is less than the second threshold, determine that the product recommendation index is the third recommendation level; among them, the third recommendation level is lower than the second recommendation level.
[0115] S211. For each group of product information, if the recommendation level of the product information is greater than a preset level threshold, input the product information into the product matching model to obtain recommended products.
[0116] Among them, the product information is screened by the preset level threshold. When the product information is less than the preset level threshold, it indicates that this product information has no recommendation value and it is not worth using this product information to generate recommended products.
[0117] For example, the recommendation level of the product information is level 3, the preset level threshold is level 2, and the recommendation level of the product information is greater than the preset level threshold. It is determined that this product information has the value of being pushed, and it is input into the product matching model to generate recommended products.
[0118] Or, the preset level threshold is level 1. Product information with recommendation levels of 2 and 3 all has the value of being recommended, but the product information with a recommendation level of 3 has a greater recommendation value. More recommended products can be generated based on the product information with a recommendation level of 3 than the product information with a recommendation level of 2.
[0119] S212. If the recommendation level of the product information is less than the preset level threshold, eliminate the product information.
[0120] Among them, if the recommended level of a product is less than the preset level threshold, it indicates that the product information has no recommendation value, and it will be excluded and no longer input into the product matching model for generating recommended products.
[0121] In this embodiment, from collecting user behavior data to generating the final product recommendations, through refined data processing, calculation, and comparison, more personalized product recommendations that meet user needs are provided, improving the accuracy of product recommendations, reducing the deviation of product recommendations, and enhancing the user's shopping experience.
[0122] Figure 3 is a schematic structural diagram of a product recommendation device provided by an embodiment of the present application. Refer to Figure 3 This product recommendation device includes various functional modules for implementing the foregoing product recommendation method, and any functional module can be implemented in software and / or hardware.
[0123] In some embodiments, the product recommendation device 300 includes a data acquisition module 301, a data parsing module 302, a recommendation index module 303, an index classification module 304, and a product matching module 305. Among them:
[0124] The data acquisition module 301 is used to acquire the shopping data of the target user; among them, the shopping data includes browsing records and purchase records;
[0125] The data parsing module 302 is used to parse the shopping data to obtain multiple groups of product information; among them, the product information includes multiple search product information and purchase product information belonging to the same product category;
[0126] The recommendation index module 303 is used to parse each group of product information based on the recommendation index algorithm to obtain the product recommendation index of each group of product information;
[0127] The index classification module 304 is used to classify each product recommendation index to obtain the recommended level of each group of product information;
[0128] The product matching module 305 is used for each group of product information. If the recommended level of the product information is greater than the preset level threshold, the product information is input into the product matching model to obtain recommended products.
[0129] In some embodiments, the data parsing module 302 is specifically used for:
[0130] Based on the browsing records, obtain multiple search product information; among them, the search product information includes the search product information category, the search product information price, and the product browsing time;
[0131] Obtain multiple purchased product information based on purchase records; among them, the purchased product information includes the category of purchased product information and the price of purchased product information;
[0132] Screen and group the search product information and purchased product information belonging to the same product category to obtain multiple groups of product information.
[0133] In some embodiments, the device further includes an information screening module 306, and the information screening module 306 is specifically configured to:
[0134] For each piece of search product information, if the product browsing time of the search product information is not greater than the preset duration threshold, then exclude the search product information.
[0135] In some embodiments, the recommendation index algorithm satisfies the following formula:
[0136]
[0137] Among them, Jzs represents the product recommendation index, SPsjz represents the product actual index, SPckz represents the product reference index, a represents the weight coefficient of the difference between the product reference index and the product actual index, and 0 < a < 1; CZm represents the time difference index of the product, b represents the weight coefficient of the time difference index, and 0 < b < 1; C represents the calibration index.
[0138] In some embodiments, the product reference index satisfies the following formula:
[0139]
[0140] Among them, SPckz represents the product reference index, SSn represents the number of search product information in the product information, JG1, JG2, JG3,..., JGn represent the prices of each piece of search product information in the product information, and D represents the first correction constant;
[0141] The product actual index satisfies the following formula:
[0142]
[0143] Among them, SPsjz represents the product actual index, GMn represents the number of purchased product information in the product information, GMJG1, GMJG2, GMJG3,..., GMJGn represent the prices of each piece of purchased product information in the product information, and E represents the second correction constant;
[0144] The time difference index satisfies the following formula:
[0145]
[0146] Wherein, CZm represents the time difference index, F represents the third correction constant, TLm represents the browsing time for searching commodity information, r represents the attenuation value, and t represents the time interval from the start of searching for commodity information to the appearance of search results.
[0147] In some embodiments, the index grading module 304 is specifically configured to:
[0148] For each commodity recommendation index, compare the commodity recommendation index with a preset first threshold and a second threshold; wherein, the first threshold is greater than the second threshold;
[0149] If the commodity recommendation index is greater than the first threshold, determine that the commodity recommendation index is the first recommendation level;
[0150] If the commodity recommendation index is less than the first threshold and the commodity recommendation index is greater than the second threshold, determine that the commodity recommendation index is the second recommendation level; wherein, the second recommendation index is lower than the first recommendation level.
[0151] In some embodiments, the index grading module 304 is specifically configured to:
[0152] If the commodity recommendation index is less than the second threshold, determine that the commodity recommendation index is the third recommendation level; wherein, the third recommendation level is lower than the second recommendation level.
[0153] The commodity recommendation device 300 provided by the embodiments of the present application is used to execute the technical solutions provided by the foregoing embodiments of the commodity recommendation method. Its implementation principle and technical effects are similar to those in the foregoing method embodiments, and will not be elaborated herein.
[0154] It should be noted that it should be understood that the division of each module of the above device is only a logical function division. In actual implementation, it can be fully or partially integrated into a physical entity, or physically separated. And these modules can all be implemented in the form of software called by processing elements, or all be implemented in the form of hardware, or some modules can be implemented in the form of software called by processing elements, and some modules can be implemented in the form of hardware. For example, the data acquisition module 301 can be a separately established processing element, or can be integrated in a certain chip of the above device. In addition, it can also be stored in the memory of the above device in the form of program code, and called and executed by a certain processing element of the above device to perform the functions of the above data acquisition module 301. The implementation of other modules is similar thereto. In addition, these modules can be fully or partially integrated together, or independently implemented. Here, the processing element can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed by the integrated logic circuit in the processor element or the instruction in the form of software.
[0155] Figure 4A schematic structural diagram of an electronic device provided by an embodiment of the present application. Refer to Figure 4 , the electronic device 400 includes: a processor 401, and a memory 402 communicatively connected to the processor 401;
[0156] The memory 402 stores computer-executable instructions;
[0157] The processor 401 executes the computer-executable instructions stored in the memory 402 to implement the technical solution of the foregoing product recommendation method.
[0158] In the above electronic device 400, the memory 402 and the processor 401 are directly or indirectly electrically connected to achieve data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines, such as being connected through a bus. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc., but it does not mean that there is only one bus or one type of bus. The memory 402 stores computer-executable instructions for implementing the foregoing product recommendation method, including at least one software function module that can be stored in the memory 402 in the form of software or firmware. The processor 401 executes various functional applications and data processing by running the software programs and modules stored in the memory 402.
[0159] The memory 402 includes at least one type of readable storage medium, which is not limited to random access memory (RAM), read only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc. Among them, the memory 402 is used to store programs. After receiving the execution instruction, the processor 401 executes the program. Further, the software programs and modules in the memory 402 may also include an operating system, which may include various software components and / or drivers for managing system tasks (such as memory management, storage device control, power management, etc.), and may communicate with various hardware or software components to provide a running environment for other software components.
[0160] The processor 401 can be an integrated circuit chip with signal processing capabilities. The above-mentioned processor 401 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a digital signal processor (DSP), an application specific integrated circuit (ASIC), etc. It can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor can be a microprocessor, or the processor 401 can also be any conventional processor, etc.
[0161] The electronic device 400 is used to execute the technical solution provided by the foregoing embodiment of the product recommendation method. Its implementation principle and technical effect are similar to those in the foregoing method embodiment, and will not be elaborated here.
[0162] The embodiments of the present application also provide a computer-readable storage medium. Computer-executable instructions are stored in the computer-readable storage medium. When the processor executes the computer-executable instructions, the technical solution of the foregoing product recommendation method is implemented.
[0163] The above computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, a magnetic disk or an optical disk. The computer-readable storage medium can be any available medium accessible by a general-purpose or special-purpose computer.
[0164] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the readable storage medium can also exist as discrete components in the control device of the merchandise recommendation device.
[0165] The embodiments of the present application also provide a computer program product, including a computer program which, when executed by a processor, is used to implement the technical solutions of the foregoing merchandise recommendation method.
[0166] In the above embodiments, those skilled in the art can understand that implementing the above method embodiments can be fully or partially realized by software, hardware, firmware, or any combination thereof. When implemented using software, it can be fully or partially realized in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions according to the embodiments of the present invention are fully or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wirelessly (such as infrared, wireless network, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server, data center, etc. that includes one or more integrated available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk (SSD)), etc.
[0167] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as falling within the scope described in this specification.
[0168] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include known common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and examples are only regarded as exemplary, and the true scope and spirit of the present application are pointed out by the following claims.
[0169] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.
Claims
1. A commodity recommendation method, characterized in that, include: Acquire the target user's shopping data; wherein the shopping data includes browsing records and purchase records; Parsing the shopping data to obtain multiple sets of product information; wherein the product information includes multiple search product information and purchase product information belonging to the same product category; Analyze each group of product information based on the recommendation index algorithm to obtain the product recommendation index of each group of product information; Classify the recommendation index of each product to obtain the recommendation level of each group of product information; For each group of product information, if the recommendation level of the product information is greater than a preset level threshold, the product information is input into a product matching model to obtain recommended products.
2. The method according to claim 1, characterized in that, The shopping data is parsed to obtain multiple sets of product information, including: Based on the browsing history, a plurality of search product information is obtained; wherein the search product information includes the search product information category, the search product information price and the product browsing time; Based on the purchase record, a plurality of purchased commodity information is obtained; wherein the purchased commodity information includes the purchased commodity information category and the purchased commodity information price; The search product information and purchase product information belonging to the same product category are filtered and grouped to obtain multiple groups of product information.
3. The method according to claim 2, wherein The method further comprises: For each searched product information, if the product browsing time of the searched product information is not greater than a preset time threshold, the searched product information is eliminated.
4. The method according to claim 1, wherein The method further comprises: The recommendation index algorithm satisfies the following formula: Where Jzs represents the recommended index of the product, SPsjz represents the actual index of the product, SPckz represents the reference index of the product, a represents the weight coefficient of the difference between the reference index of the product and the actual index of the product, and 0 <a<1;CZm表示商品的时间差值指数,b表示时间差值指数的权重系数,且0<b<1;C表示校准指数。 5. The method according to claim 4, wherein The method further comprises: The commodity reference index satisfies the following formula: Wherein, SPckz represents the commodity reference index, SSn represents the number of search commodity information in the commodity information, JG1, JG2, JG3, ..., JGn represent the price of each search commodity information in the commodity information, and D represents the first correction constant; The actual index of the commodity satisfies the following formula: Wherein, SPsjz represents the actual index of the commodity, GMn represents the number of purchased commodity information in the commodity information, GMJG1, GMJG2, GMJG3, ..., GMJGn represent the price of each purchased commodity information in the commodity information, and E represents the second correction constant; The time difference index satisfies the following formula: Wherein, CZm represents the time difference index, F represents the third correction constant, TLm represents the browsing time of searching for product information, r represents the decay value, and t represents the time interval from the start of searching for product information to the appearance of search results.
6. The method according to claim 1, characterized in that The recommendation index of each product is graded to obtain the recommendation level of each group of product information, including: For each product recommendation index, compare the product recommendation index with a preset first threshold and a second threshold; wherein the first threshold is greater than the second threshold; If the product recommendation index is greater than the first threshold, determine that the product recommendation index is the first recommendation level; If the product recommendation index is less than the first threshold and the product recommendation index is greater than the second threshold, determine that the product recommendation index is the second recommendation level; wherein, the second recommendation index is lower than the first recommendation level; If the product recommendation index is less than the second threshold, determine that the product recommendation index is the third recommendation level; wherein, the third recommendation level is lower than the second recommendation level.
7. The method according to any one of claims 1-6, characterized in that, The method further includes: If the recommendation level of the product information is less than the preset level threshold, eliminate the product information.
8. A commodity recommendation device, characterized in that, It includes: A data acquisition module, configured to acquire the shopping data of a target user; wherein, the shopping data includes browsing records and purchase records; A data parsing module, configured to parse the shopping data to obtain multiple groups of product information; wherein, the product information includes multiple search product information and purchase product information belonging to the same product category; A recommendation index module, configured to parse each group of product information based on a recommendation index algorithm to obtain the product recommendation index of each group of product information; An index grading module, configured to grade each product recommendation index to obtain the recommendation level of each group of product information; A product matching module, configured to, for each group of product information, if the recommendation level of the product information is greater than the preset level threshold, input the product information into a product matching model to obtain recommended products.
9. An electronic device, characterized in that, It includes: A processor, and a memory communicatively connected to the processor; The memory stores computer execution instructions; The processor executes the computer execution instructions stored in the memory to implement the method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, Computer execution instructions are stored in the computer-readable storage medium, and when the computer execution instructions are executed by a processor, they are used to implement the method according to any one of claims 1 to 7.