Product pushing method, device and equipment and storage medium
By analyzing user data to calculate product and cover preferences, personalized product covers are constructed, solving the problem of inaccurate product push in existing technologies and improving user experience and push efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- PING AN BANK CO LTD
- Filing Date
- 2022-08-22
- Publication Date
- 2026-07-07
Smart Images
Figure CN115329199B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of artificial intelligence, and more particularly to a product push method, apparatus, device, and storage medium. Background Technology
[0002] With the rapid development of computer technology, more and more online promotion methods have emerged. However, when business personnel push product data online, they cannot effectively push highly relevant product data based on limited user information.
[0003] Traditional product data push methods typically involve directly sending product data to user terminals based on existing user information, failing to achieve personalized recommendations. Consequently, the products pushed to users often fail to stimulate their purchasing desire, and users struggle to find products of interest from the vast sea of recommendations. Therefore, users increasingly demand personalized product recommendations, with the expectation that requesting the product during the recommendation process can further enhance their purchase interest. Thus, how to achieve personalized product recommendations and further improve the user experience is a pressing issue that needs to be addressed. Summary of the Invention
[0004] In view of this, in order to solve the problems of the prior art, the present invention provides a product push method, apparatus, device and storage medium that can be applied to fields such as fintech or other fields.
[0005] In a first aspect, the present invention provides a product push method, comprising:
[0006] Obtain user registration information, historical transaction records, and product reviews;
[0007] The registration information, historical transaction records, and product reviews are analyzed to obtain the preference for each product and the preference for product cover images.
[0008] Based on the product cover preference, analyze the user's preferred style of each product cover to obtain the user's cover style preference;
[0009] Sort the products according to the product preference, and select a preset number of preferred products;
[0010] Based on the aforementioned cover style preferences, construct the product cover for the preferred product;
[0011] The preferred products, which include the product cover, will be pushed to the user's device.
[0012] In an optional implementation, the analysis of the registration information, the historical transaction records, and the product reviews to obtain individual product preferences and product cover preferences includes:
[0013] Based on the registration information, calculate the first product preference and the first product cover preference;
[0014] Based on the historical transaction records, calculate the purchase rate of each product, and determine the second product preference and the second product cover preference based on the purchase rate;
[0015] Sentiment analysis was performed on the product reviews to determine the third product preference and the third product cover preference.
[0016] Based on the first product preference, the second product preference, and the third product preference, the user's product preference for each product is determined;
[0017] Based on the first product cover preference, the second product cover preference, and the third product cover preference, the user's product cover preference for each product cover is determined.
[0018] In an optional implementation, the step of performing sentiment analysis on the product evaluation to determine the third product preference and the third product cover preference includes:
[0019] The product evaluations are categorized to obtain first product evaluation data containing the product and second product evaluation data containing the product cover.
[0020] The first product evaluation data and the second product evaluation data are respectively processed by word segmentation to obtain the first word to be analyzed and the second word to be analyzed.
[0021] The first word to be analyzed and the second word to be analyzed are matched with sentiment words in the preset word library and the corresponding first sentiment tendency value and second sentiment tendency value are calculated.
[0022] The preference for the third product is determined based on the first sentiment tendency value, and the preference for the third product cover is determined based on the second sentiment tendency value.
[0023] In an optional implementation, each sentiment word in the preset lexicon corresponds to a sentiment reference value. The step of matching the first and second words to be analyzed with sentiment words in the preset lexicon and calculating the corresponding first and second sentiment inclination values includes:
[0024] The first word to be analyzed and the second word to be analyzed are matched with the sentiment words in the preset word library to obtain the corresponding target sentiment words; wherein, a product or a product cover corresponds to at least one of the target sentiment words;
[0025] Calculate the product of the preset weight and the emotional reference value corresponding to the target emotional word to obtain the first emotional tendency value of each product and the second emotional tendency value of each product cover.
[0026] In an optional implementation, determining the preference for each product cover based on the first product cover preference, the second product cover preference, and the third product cover preference includes:
[0027] Based on preset weighting coefficients, the first product cover preference, the second product cover preference, and the third product cover preference corresponding to the same product are weighted and calculated to obtain the product cover preference of the product.
[0028] Calculate the product cover preference for each of the products.
[0029] In an optional implementation, the step of sorting the products according to the product preference and selecting a preset number of preferred products includes:
[0030] The products are categorized to obtain multiple product types;
[0031] Based on the product preference degree of each product, calculate the type preference degree corresponding to each product type;
[0032] The product types are sorted in descending order of preference for each type, and the top N product types are selected, where N is a natural number.
[0033] Select a preset number of preferred products from the N product types.
[0034] In an optional implementation, pushing the preferred product containing the product cover to the user terminal includes:
[0035] The preferred products, including the product cover, are pushed to the user's device in a randomized order, so that the user's device randomly displays the preferred products; or,
[0036] The preferred products containing the product cover are pushed to the user terminal in descending order of product preference, so that the user terminal can prioritize displaying the preferred products with higher product preference.
[0037] In a second aspect, the present invention provides a product delivery device, comprising:
[0038] The acquisition module is used to obtain users' registration information, historical transaction records, and product reviews.
[0039] The first analysis module is used to analyze the registration information, the historical transaction records, and the product reviews to obtain the preference for each product and the preference for the product cover.
[0040] The second analysis module is used to analyze the user's preferred style of each product cover based on the product cover preference, and to obtain the user's cover style preference.
[0041] The sorting module is used to sort the products according to the product preference and select a preset number of preferred products.
[0042] A construction module is used to construct the product cover of the preferred product based on the cover style preference.
[0043] The push module is used to push the preferred products containing the product cover to the user terminal.
[0044] Thirdly, the present invention provides a computer device, the computer device including a memory and at least one processor, the memory storing a computer program, and the processor executing the computer program to implement the aforementioned product delivery method.
[0045] Fourthly, the present invention provides a computer storage medium storing a computer program, which, when executed, implements the aforementioned product push method.
[0046] The embodiments of the present invention have the following beneficial effects:
[0047] This embodiment analyzes user registration information, historical transaction records, and product reviews to calculate user preferences for various products and product cover images. Based on these preferences, products are pushed to users, achieving personalized recommendations. The pushed products are those that the user is interested in or prefers, improving product push efficiency. Furthermore, when pushing products, the system constructs user-preferred product covers based on the user's cover image preferences, pushing selected products containing these covers to the user, improving user experience and satisfaction, and ultimately enhancing system vitality. Attached Figure Description
[0048] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope of protection of the present invention. In the various drawings, similar components are numbered similarly.
[0049] Figure 1 This diagram illustrates a first embodiment of the product push method according to the present invention.
[0050] Figure 2 A schematic diagram of a second embodiment of the product push method in this invention is shown;
[0051] Figure 3 A schematic diagram of a third embodiment of the product push method in this invention is shown;
[0052] Figure 4 A schematic diagram of the fourth embodiment of the product push method in this invention is shown;
[0053] Figure 5 A schematic diagram of the fifth embodiment of the product push method in this invention is shown;
[0054] Figure 6 A schematic diagram of the sixth embodiment of the product push method in this invention is shown;
[0055] Figure 7 A schematic diagram of the product pushing device in an embodiment of the present invention is shown. Detailed Implementation
[0056] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.
[0057] The components of the embodiments of the invention described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.
[0058] In the following, the terms “comprising,” “having,” and their cognates, which may be used in various embodiments of the invention, are intended only to indicate a particular feature, number, step, operation, element, component, or combination thereof, and should not be construed as excluding, firstly, the presence of one or more other features, numbers, steps, operations, elements, components, or combinations thereof, or adding the possibility of one or more features, numbers, steps, operations, elements, components, or combinations thereof.
[0059] Furthermore, the terms "first," "second," and "third" are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.
[0060] Unless otherwise specified, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of the invention pertain. Terms (such as those defined in commonly used dictionaries) shall be interpreted as having the same meaning as in their contextual meaning in the relevant technical field and shall not be interpreted as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of the invention.
[0061] Example 1
[0062] In real-world scenarios, businesses can promote their products to users online. For example, during banking transactions, each bank pushes various products, including wealth management products and promotional items, to users through its website or application. However, existing user data mining or product data-driven push methods cannot personalize the push to each user to meet their needs or interests, resulting in low product push efficiency. Furthermore, overly simplistic product images can easily fail to attract user interest and generate purchase desire. Therefore, this embodiment provides a product push method that enables personalized push of products and their images, improving user experience and product push efficiency.
[0063] Please refer to Figure 1 The following is a detailed explanation of the product push method.
[0064] S10 obtains user registration information, historical transaction records, and product reviews.
[0065] The server obtains the user's registration information during the account registration process, as well as the user's historical transaction records and product reviews during account usage.
[0066] As an example, users are required to register when they first log in to the bank's website or application. During the registration process, the system pre-sets multiple product (or product type) and / or product cover (or product cover style) options for users to choose from. Users can select or enter their own preferred products and / or product covers (or product types and / or product cover styles); of course, users can also skip this step and register directly. Therefore, the user's registration information includes personal account information and data such as the products or product covers the user selected or entered. When users log in to the bank's website or application and purchase or rate corresponding products or product covers through their accounts, the backend automatically records their purchase and rating records, as well as the specific purchase details and rating content.
[0067] S20 analyzes registration information, historical transaction records, and product reviews to obtain the preference for each product and the preference for product cover.
[0068] Natural language processing is performed on registration information, historical transaction records, and product reviews to extract the language descriptions of the products and product covers involved in the registration information, historical transaction records, and product reviews. The language descriptions are then analyzed to calculate the user's preference for each product and the product cover.
[0069] As one possible implementation method, please refer to Figure 2 Step S20 may specifically include the following steps:
[0070] S21, Calculate the first product preference and the first product cover preference based on the registration information.
[0071] Based on the products and / or product covers (or product types and / or product cover styles) of interest selected or entered by the user in their registration information, calculate the user's first product preference and first product cover preference for each product. Furthermore, a predetermined higher preference level is set for all products and product covers selected by the user. Additionally, a preference level of zero or a predetermined lower preference level can be set for other products and product covers not selected by the user. The predetermined preference levels can be set according to actual circumstances and are not limited here.
[0072] S22, calculate the purchase rate of each product based on historical transaction records, and determine the preference of the second product and the preference of the second product cover based on the purchase rate.
[0073] Extract the user's purchase information for each product from the user's historical transaction records, calculate the quantity of each product purchased by the user, calculate the purchase rate of each product, and determine the user's product preference for each product and the product cover preference for that product based on the purchase rate.
[0074] Furthermore, the purchase history can be used to identify the product types of each purchased product and calculate the purchase rate of each product type. The purchase rate of each product type can then be used to determine the product types that users are interested in.
[0075] S23, Conduct sentiment analysis on product evaluations to determine the preference for the third product and the preference for the third product cover.
[0076] Obtain at least one sentence containing a product or product cover from product reviews. Perform semantic analysis on these sentences, and then perform sentiment analysis on the product or product cover mentioned in the sentence based on the semantics of the sentence. This is to determine the user's product preference and product cover preference for each product based on the product reviews. Sentiment preference includes positive, negative, and neutral sentiment.
[0077] This embodiment uses sentiment analysis of product reviews to more accurately determine users' preferences for various products or product covers, improving the accuracy of calculating product or product cover preferences. This further enhances the reliability and accuracy of personalized recommendations based on users' preferences for various products or product covers.
[0078] As one possible implementation method, please refer to Figure 3 Step S23 may specifically include the following steps:
[0079] S231, classify the product evaluations to obtain first product evaluation data containing the product and second product evaluation data containing the product cover.
[0080] Natural language processing is used to extract statements containing keywords related to a specific product, which are then used as product reviews for that product. Similarly, statements containing keywords related to a product's cover image are extracted and used as cover image reviews. This allows for the extraction of review statements for each product or product cover image from all of a user's product reviews. In other words, product reviews are categorized based on the keywords they contain, resulting in first-level product review data containing the product itself and second-level product review data containing the product cover image.
[0081] S232, perform word segmentation on the first product evaluation data and the second product evaluation data respectively to obtain the first word to be analyzed and the second word to be analyzed.
[0082] The first and second product evaluation data are processed by word segmentation, that is, the first product evaluation statement containing the product and the second product evaluation statement containing the product cover are processed by word segmentation, resulting in the first and second words to be analyzed. The word segmentation process can be implemented using a pre-defined word segmentation tool or algorithm with word segmentation functionality; the specific word segmentation tool or algorithm is not limited here.
[0083] S233, match the first and second words to be analyzed with the sentiment words in the preset word library and calculate the corresponding first sentiment tendency value and second sentiment tendency value.
[0084] The first and second words to be analyzed are matched with sentiment words in the preset word library to calculate the user's first sentiment tendency value for each product and the user's second sentiment tendency value for each product cover.
[0085] Specifically, the sentiment tendency value can be determined based on the number of sentiment words that match the first and second words to be analyzed. For example, if the number of words matching positive sentiment words in the first word to be analyzed is greater than the number matching negative sentiment words, and the number of words matching positive sentiment words is significantly higher than the number matching negative sentiment words, then the sentiment tendency corresponding to the first word to be analyzed is positive. Therefore, a mapping relationship between the number of words matching sentiment words and the sentiment tendency value can be established to determine the corresponding sentiment tendency value. The setting of this mapping relationship can be determined according to the actual situation and is not limited here.
[0086] As one possible implementation method, please refer to Figure 4 When each sentiment word in the preset lexicon corresponds to a sentiment reference value, step S233 specifically includes the following steps:
[0087] S2331, Match the first word to be analyzed and the second word to be analyzed with the sentiment words in the preset word library respectively to obtain the target sentiment words; wherein, a product or a product cover corresponds to at least one target sentiment word.
[0088] The first and second words to be analyzed are matched with sentiment words in the preset word library that represent positive, negative, and neutral tendencies, respectively, to obtain the corresponding target sentiment words. Among them, a product or a product cover corresponds to at least one target sentiment word.
[0089] S2332, calculate the product of the preset weight and the emotional reference value corresponding to the target emotional words, and obtain the first emotional tendency value of each product and the second emotional tendency value of each product cover.
[0090] Based on the relevance of sentiment words to the bank's products or product covers, different sentiment reference values are assigned to different sentiment words. For example, positive sentiment words are assigned positive reference values, with higher values indicating stronger positive emotions; negative sentiment words are assigned negative reference values, with lower values indicating stronger negative emotions; and neutral sentiment words are assigned zero sentiment reference values. It is understood that the specific sentiment reference values can be set according to actual circumstances and are not limited here.
[0091] For a product or product cover, the emotional tendency sub-value is obtained by calculating the product of preset weights and the emotional reference values corresponding to the target emotional words. Specifically, the same or different weights are set for positive, negative, and neutral tendencies, and the emotional tendency sub-value of the product or product cover is calculated based on the emotional reference values of each emotional word corresponding to the product or product cover and their corresponding weights. Similarly, the emotional tendency value corresponding to each product or product cover can be obtained. The specific values of the preset weights are not limited here and can be set according to actual needs.
[0092] S234, determine the preference for the third product based on the first sentiment tendency value, and determine the preference for the third product cover based on the second sentiment tendency value.
[0093] The user's preference for a product or its cover is determined based on the sentiment index corresponding to each product or product cover. For example, a positive sentiment index indicates a positive preference for the product or cover, and the higher the sentiment index, the higher the preference. Furthermore, a mapping relationship between the magnitude of the sentiment index and the preference can be pre-set to determine the user's preference for each product or product cover. The specific mapping relationship between the magnitude of the sentiment index and the preference can be set according to actual needs and is not limited here.
[0094] S24. Determine the user's product preference for each product based on the first product preference, the second product preference, and the third product preference.
[0095] By comprehensively calculating the first, second, and third product preferences for the same product, the user's product preference for that product can be obtained. Similarly, the user's product preference for all products can be calculated.
[0096] S25, determine the user's product cover preference for each product cover based on the first product cover preference, the second product cover preference, and the third product cover preference.
[0097] The user's preference for the first, second, and third product covers corresponding to the same product cover is calculated by combining these preferences. Similarly, the user's preference for all product covers can be calculated.
[0098] As one possible implementation method, please refer to Figure 5 Step S25 may specifically include the following steps:
[0099] S251, based on preset weighting coefficients, calculate the weights of the first product cover preference, the second product cover preference, and the third product cover preference corresponding to the same product to obtain the product cover preference.
[0100] The product cover preference scores for the same product (first, second, and third product covers) are weighted according to preset weighting coefficients to calculate the overall product cover preference score. The weighting values for user registration information, historical transaction records, and product reviews can be preset to calculate the overall product cover preference score based on these three factors. The preset weighting coefficients or values can be set according to actual circumstances and are not limited here.
[0101] S252, calculate the product cover preference for each product.
[0102] Once the product cover preference for the same product is obtained, the product cover preference for all products can be calculated, which means calculating the product cover preference for each individual product.
[0103] S30: Based on product cover preference, analyze the user's preferred style of each product cover to obtain the user's cover style preference.
[0104] Based on the product cover preference for each product, analyze the style of each product cover preferred by users; that is, analyze the degree of user preference for the type of each product cover.
[0105] Furthermore, the product covers are first categorized according to their corresponding styles. This categorization can be based on the style characteristics of each product cover, grouping covers belonging to the same style characteristic into the same style type. Then, the preference score for that style type is calculated based on the preference scores of the product covers within that style type. Specifically, each style type corresponds to at least one style characteristic; product covers containing the corresponding style characteristic are grouped into one style type, and each style type includes at least one product cover. The preference scores of the product covers within a style type are then added together or calculated using a predetermined coefficient to obtain the user's cover style preference.
[0106] S40: Sort the products according to product preference and select a preset number of preferred products.
[0107] Each product is sorted according to its corresponding product preference level, and a preset number of preferred products are selected. The preferred products are those with higher product preference levels, and the preset number can be set according to actual conditions, without limitation here.
[0108] As one possible implementation method, please refer to Figure 6 Step S40 further includes the following steps:
[0109] S41, classify the products to obtain multiple product types.
[0110] The products are categorized to obtain multiple product types, with each product type corresponding to at least one product. Further, the categorization process can be based on at least one product characteristic (or feature) shared by each product, grouping products containing one or more of the same characteristic into a single product type. In other words, each product type corresponds to at least one product characteristic, and products containing the corresponding characteristic are grouped into one product type. A product type may include at least one product.
[0111] S42, Calculate the type preference degree corresponding to each product type based on the product preference degree of each product.
[0112] Based on the product preference and product type for each product, the type preference for each product type is calculated. Specifically, the preference values of each product within a product type are added together or calculated using a predetermined coefficient to obtain the type preference for each product type.
[0113] S43. Sort each product type in descending order of type preference, and select the top N product types, where N is a natural number.
[0114] From multiple product types, product types with higher type preference are selected to facilitate the selection of multiple products as preferred products. Specifically, each product type is sorted in descending order of type preference, and then the top N product types are selected, where N is a natural number.
[0115] S44: Select a preset number of preferred products from N product types.
[0116] Select a preset number of preferred products from N product types. The preset number can be set according to the actual situation and is not limited here.
[0117] For example, select X products from N product types, that is, select X products as preferred products from each of the N product types, which means selecting (N*X) preferred products, where X is a natural number; or, arbitrarily select a certain number of products as preferred products from each of the N product types, and the number of preferred products selected in each product type is not necessarily equal.
[0118] S50: Based on cover style preferences, construct product covers for preferred products.
[0119] Based on different users' cover style preferences, product covers for preferred products are constructed. In this process, the cover style with the highest preference is selected as the design style for the product cover. Then, an intelligent image generator is used to automatically generate the product cover for the preferred product based on the preferred product and the cover style.
[0120] As an example, this intelligent image generator can utilize Dall-E (an image generation system), which can create extremely realistic and clear images based on simple descriptions, proficient in various art styles, including illustration and landscapes. It can also generate text for building signage and create sketches and full-color images of the same scene. Furthermore, Dall-E can be subjected to deep learning to generate product cover images that meet specific requirements.
[0121] S60 pushes selected products, including product covers, to the user's device.
[0122] The selected products, including their covers, are pushed to the user's device. In other words, the selected products and their covers are pushed to the user.
[0123] In one implementation, when pushing products to users, a selection of preferred products, including product covers, can be pushed to the user's device in a random order, so that the user's device displays the preferred products randomly. That is, when the user's device receives the preferred products, the display effect is consistent, the display order is random, there is no situation where a particular product is highlighted, and the user can randomly select products that interest them to learn about or purchase.
[0124] In one implementation, when pushing products to a user, preferred products, including product covers, are pushed to the user's device in descending order of product preference, so that the user's device prioritizes displaying the preferred products with the highest product preference. Under this push method, the user's device can set different or the same display effects for the received preferred products, but the display order will be based on the preference level of each preferred product. For example, if there are M preferred products, the M preferred products are sorted in descending order of their respective preference levels, and the top N preferred products are selected for priority display, where N and M are both positive integers, and N is less than M. Further, the priority display method can be that the selected N preferred products are displayed first on the product display interface, then other preferred products are displayed in pages, and finally, other products are displayed. That is, the user can first view the selected N preferred products on the product display page, then click the next page, scroll down the page, or turn the page to view other preferred products, and finally view other products on the next page.
[0125] In this embodiment, by analyzing user registration information, historical transaction records, and product reviews, the user's preference for the product and its cover is calculated. Based on this preference, a product cover with the user's preferred cover style is constructed, and preferred products containing this constructed cover are pushed to the user. This embodiment achieves personalized recommendations for users, ensuring that the products pushed to users are those that the user is interested in or prefers, thus improving product push efficiency. Furthermore, when pushing products, the system also constructs a user-preferred product cover based on the user's cover preferences, pushing preferred products containing this constructed cover to the user, improving user experience and satisfaction, and ultimately enhancing system vitality.
[0126] Example 2
[0127] Please see Figure 7 This invention provides a product delivery device, comprising:
[0128] Module 71 is used to obtain users' registration information, historical transaction records, and product reviews;
[0129] The first analysis module 72 is used to analyze the registration information, the historical transaction records and the product evaluations to obtain the preference for each product and the preference for the product cover.
[0130] The second analysis module 73 is used to analyze the user's preferred style of each product cover based on the product cover preference, and obtain the user's cover style preference.
[0131] The sorting module 74 is used to sort the products according to the product preference and select a preset number of preferred products.
[0132] Module 75 is used to construct the product cover of the preferred product based on the cover style preference.
[0133] The push module 76 is used to push the preferred product containing the product cover to the user terminal.
[0134] The product push device described above corresponds to the product push method in Embodiment 1. Any option in Embodiment 1 is also applicable to this embodiment, and will not be described in detail here.
[0135] This invention also provides a computer device, which includes a memory and at least one processor. The memory stores a computer program, and the processor executes the computer program to implement the product delivery method described above.
[0136] The memory may include a stored program area and a stored data area. The stored program area may store the operating system and at least one application program required for a function. The stored data area may store data created based on the use of the computer device (such as preferred products, product preferences, etc.). In addition, the memory may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.
[0137] This invention also provides a computer-readable storage medium storing machine-executable instructions. When called and executed by a processor, the machine-executable instructions cause the processor to perform the steps of the product delivery method described above.
[0138] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that, as an alternative implementation, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.
[0139] In addition, the functional modules or units in the various embodiments of the present invention can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.
[0140] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a smartphone, personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0141] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.
Claims
1. A product push method, characterized in that, include: Obtain user registration information, historical transaction records, and product reviews; Based on the registration information, calculate the first product preference and the first product cover preference; Based on the historical transaction records, calculate the purchase rate of each product, and determine the second product preference and the second product cover preference based on the purchase rate; The product reviews are categorized to obtain first product review data containing the product and second product review data containing the product cover; the first product review data and the second product review data are segmented into words to obtain a first word to be analyzed and a second word to be analyzed; the first word to be analyzed and the second word to be analyzed are matched with sentiment words in a preset word library and the corresponding first sentiment tendency value and second sentiment tendency value are calculated. The preference for the third product is determined based on the first sentiment tendency value, and the preference for the third product cover is determined based on the second sentiment tendency value. Based on the first product preference, the second product preference, and the third product preference, the user's product preference for each product is determined; Based on the first product cover preference, the second product cover preference, and the third product cover preference, the user's product cover preference for each product cover is determined; Based on the product cover preference, analyze the user's preferred style of each product cover to obtain the user's cover style preference; Sort the products according to the product preference, and select a preset number of preferred products; Based on the aforementioned cover style preferences, construct the product cover for the preferred product; The preferred products, which include the product cover, will be pushed to the user's device.
2. The product push method according to claim 1, characterized in that, Each sentiment word in the preset lexicon corresponds to a sentiment reference value. The process of matching the first and second words to be analyzed with sentiment words in the preset lexicon and calculating the corresponding first and second sentiment inclination values includes: The first word to be analyzed and the second word to be analyzed are matched with the sentiment words in the preset word library to obtain the corresponding target sentiment words; wherein, a product or a product cover corresponds to at least one of the target sentiment words; Calculate the product of the preset weight and the emotional reference value corresponding to the target emotional word to obtain the first emotional tendency value of each product and the second emotional tendency value of each product cover.
3. The product push method according to claim 1, characterized in that, The step of determining the preference for each product cover based on the first product cover preference, the second product cover preference, and the third product cover preference includes: Based on preset weighting coefficients, the product cover preference of the first product cover, the second product cover preference, and the third product cover preference corresponding to the same product are weighted and calculated to obtain the product cover preference of the product. Calculate the product cover preference for each of the products.
4. The product push method according to claim 1, characterized in that, The step of sorting the products according to the product preference and selecting a preset number of preferred products includes: The products are categorized to obtain multiple product types; Based on the product preference degree of each product, calculate the type preference degree corresponding to each product type; The product types are sorted in descending order of preference for each type, and the top N product types are selected, where N is a natural number. Select a preset number of preferred products from the N product types.
5. The product push method according to claim 1, characterized in that, The step of pushing the preferred product containing the product cover to the user includes: The preferred products, including the product cover, are pushed to the user's device in a randomized order, so that the user's device randomly displays the preferred products; or, The preferred products containing the product cover are pushed to the user terminal in descending order of product preference, so that the user terminal can prioritize displaying the preferred products with higher product preference.
6. A product pushing device, characterized in that, include: The acquisition module is used to obtain users' registration information, historical transaction records, and product reviews. The first analysis module is used to calculate the first product preference degree and the first product cover preference degree based on the registration information; Based on the historical transaction records, calculate the purchase rate of each product, and determine the second product preference and the second product cover preference based on the purchase rate; The product reviews are categorized to obtain first product review data containing the product and second product review data containing the product cover; the first product review data and the second product review data are segmented into words to obtain a first word to be analyzed and a second word to be analyzed; the first word to be analyzed and the second word to be analyzed are matched with sentiment words in a preset word library and the corresponding first sentiment tendency value and second sentiment tendency value are calculated. The preference for the third product is determined based on the first sentiment tendency value, and the preference for the third product cover is determined based on the second sentiment tendency value. Based on the first product preference, the second product preference, and the third product preference, the user's product preference for each product is determined; Based on the first product cover preference, the second product cover preference, and the third product cover preference, the user's product cover preference for each product cover is determined; The second analysis module is used to analyze the user's preferred style of each product cover based on the product cover preference, and to obtain the user's cover style preference. The sorting module is used to sort the products according to the product preference and select a preset number of preferred products. A construction module is used to construct the product cover of the preferred product based on the cover style preference. The push module is used to push the preferred products containing the product cover to the user terminal.
7. A computer device, characterized in that, The computer device includes a memory and at least one processor, the memory storing a computer program, and the processor executing the computer program to implement the product delivery method according to any one of claims 1-5.
8. A computer storage medium, characterized in that, It stores a computer program, which, when executed, implements the product push method according to any one of claims 1-5.