A personalized advertising push method based on product data and related device

Abstract

The present application provides a personalized advertising push method and related devices based on product data. Obtain product data of the product to be promoted, collect basic information and sales data of the product to be promoted, and provide basic data for the subsequent recommendation algorithm. By identifying similar products, more choices can be provided to users to avoid single recommendations. At the same time, collecting similar product data can help the algorithm analyze user needs more accurately. The weight of each similar product is calculated by analyzing similar product data, which reflects the relevance of each similar product to the user's interests. Sort by weight coefficient, and select similar products corresponding to the first k weight coefficients as recommended products. This ensures the relevance and effectiveness of the recommended products. By displaying advertising information, users can easily obtain the latest recommendation information, improve the reach of advertising and user experience.

Description

Technical Field

[0001] The present application relates to the technical field of Internet and e-commerce, and in particular to a personalized advertising push method based on product data and related devices. Background Art

[0002] With the rapid development of the Internet, online shopping has become an indispensable part of people's lives. However, in order to increase their sales or expand the popularity of their brands, merchants also promote their products through various channels, such as displaying them on the right side of the search results page or directly sending emails to target customers. Although these methods can make some people pay attention to a product and facilitate transactions, this method also has some disadvantages: (1) If you are not familiar with the product, you may misunderstand it after seeing the advertisement, which may lead to giving up buying the product. (2) Generally, people will only click on the link and read the relevant information when they are interested, which will result in a lot of time wasted and is not conducive to reducing costs.

[0003] The methods described in this section are not necessarily methods that have been previously conceived or employed. Unless otherwise indicated, it should not be assumed that any method described in this section is considered to be prior art simply because it is included in this section. Similarly, unless otherwise indicated, the issues mentioned in this section should not be considered to have been recognized in any prior art. Summary of the invention

[0004] In view of this, an embodiment of the present application provides a personalized advertising push method based on product data and a related device.

[0005] In a first aspect, an embodiment of the present application provides a method for personalized advertising push based on product data, the method comprising:

[0006] Acquire product data of the product to be promoted; the product data contains basic attributes of the product to be promoted and sales statistics of the product to be promoted;

[0007] Determine at least two similar products of the product to be promoted, and determine product data of each of the at least two similar products; the product to be promoted and the at least two similar products belong to the same product category;

[0008] Calculate the weight coefficients of the at least two similar products, extract the first k weight coefficients from all the data weight coefficients by data sorting, and determine the similar products corresponding to the first k weight coefficients as recommended products; k is an integer greater than or equal to 1;

[0009] The product data corresponding to the recommended product is displayed on the user's display desktop or browser page.

[0010] In a possible implementation, the calculating the weight coefficients of the at least two similar products includes:

[0011] Constructing a first feature vector according to the product data of the product to be promoted;

[0012] constructing at least two second feature vectors according to the product data of the at least two similar products;

[0013] The distance between the first feature vector and each of the second feature vectors is calculated to obtain a weight coefficient of the at least two similar products.

[0014] In a possible implementation, displaying the product data corresponding to the recommended product on the user's display desktop or browser page includes:

[0015] For each recommended product, the relevance score between the recommended product and the user is calculated based on the user's personal information and historical behavior data, and the recommended product with the highest score is displayed as the final recommended product on the user's display desktop or browser page.

[0016] In a possible implementation, the step of obtaining product data of the product to be promoted includes:

[0017] The product data of the product to be promoted is obtained by online data crawling, database query or user input.

[0018] In a possible implementation, the method further includes:

[0019] The user's response to the recommended product during the advertisement recommendation process is recorded, and the recommended product is adjusted according to the response result.

[0020] In a second aspect, an embodiment of the present application provides a personalized advertising push device based on product data, the device comprising:

[0021] An acquisition module, used to acquire product data of the product to be promoted; the product data contains basic attributes of the product to be promoted and sales statistics of the product to be promoted;

[0022] A determination module, used to determine at least two similar products of the product to be promoted, and to determine the product data of each of the at least two similar products; the product to be promoted and the at least two similar products belong to the same product category;

[0023] A calculation module, used to calculate the weight coefficients of the at least two similar products, and extract the first k weight coefficients from all the data weight coefficients by data sorting, and determine the similar products corresponding to the first k weight coefficients as recommended products; k is an integer greater than or equal to 1;

[0024] The display module is used to display the product data corresponding to the recommended product on the user's display desktop or browser page.

[0025] In a possible implementation manner, the calculation module is specifically used to:

[0026] Constructing a first feature vector according to the product data of the product to be promoted;

[0027] constructing at least two second feature vectors according to the product data of the at least two similar products;

[0028] The distance between the first feature vector and each of the second feature vectors is calculated to obtain a weight coefficient of the at least two similar products.

[0029] In a possible implementation, the display module is specifically used to:

[0030] For each recommended product, the relevance score between the recommended product and the user is calculated based on the user's personal information and historical behavior data, and the recommended product with the highest score is displayed as the final recommended product on the user's display desktop or browser page.

[0031] In a possible implementation, the acquisition module is specifically used to:

[0032] The product data of the product to be promoted is obtained by online data crawling, database query or user input.

[0033] In a possible implementation manner, the device further includes:

[0034] The adjustment module is used to record the user's response results to the recommended products during the advertisement recommendation process, and adjust the recommended products according to the response results.

[0035] In a third aspect, an embodiment of the present application provides a device, comprising a memory and a processor, wherein the memory is used to store instructions or codes, and the processor is used to execute the instructions or codes so that the device executes the personalized advertising push method based on product data as described in any one of the aforementioned first aspects.

[0036] In a fourth aspect, an embodiment of the present application provides a computer storage medium, in which a code is stored. When the code is executed, the device executing the code implements the personalized advertising push method based on product data as described in any one of the above-mentioned first aspects.

[0037] The personalized advertising push method based on product data and related devices provided by this application have the following technical effects:

[0038] 1. Accurate recommendation: By analyzing the attributes, sales data, and user data of the products to be promoted, and combining the weight coefficients of similar products, the solution can more accurately identify user needs and recommend advertisements that are highly matched with user interests, thereby improving the click-through rate and conversion rate of advertisements.

[0039] 2. Rich recommendations: By identifying similar products, the solution can provide users with more choices, meet their diverse needs, and improve advertising appeal and user satisfaction.

[0040] 3. Efficient push: By displaying product data on the electronic devices used by users, advertising information can be effectively pushed to users, thereby increasing the reach of advertising.

[0041] 4. Improve user experience: Through accurate recommendations and diversified choices, this solution can reduce users’ annoyance with advertisements, increase users’ acceptance of advertisements, and improve users’ overall browsing experience.

[0042] 5. Scalability: The solution can be easily extended to other product categories and platforms, and has high scalability and versatility. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in this embodiment or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 A flowchart of a personalized advertising push method based on product data provided in an embodiment of the present application;

[0045] Figure 2 A schematic diagram of the structure of a personalized advertising push device based on product data provided in an embodiment of the present application. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. The terms used in the following embodiments are only for the purpose of describing specific embodiments and are not intended to be used as limitations on the present application. As used in the specification and the appended claims of the present application, the singular expressions "one", "a kind", "said", "above", "the" and "this" are intended to also include expressions such as "one or more", unless there is a clear contrary indication in the context. It should also be understood that in the embodiments of the present application, "one or more" refers to one, two or more; "and / or" describes the association relationship of the associated objects, indicating that three relationships may exist; for example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship.

[0047] It should be understood that the specific embodiments described herein are merely examples of the inventive concept of the present invention and its interpretation method, and are not intended to limit the scope of the embodiments of the present invention. The present invention will be described in detail below with reference to the accompanying drawings and examples. It should be noted that, in the absence of conflict, the various embodiments in this application can be combined with each other to form other new embodiments.

[0048] like Figure 1 As shown, Figure 1 1 is a flow chart of a personalized advertising push method provided by an embodiment of the present invention. The method comprises:

[0049] S101: Acquire product data of the product to be promoted.

[0050] The product data contains the basic attributes of the product to be promoted and the sales statistics of the product to be promoted. Product data can be obtained from the internal database or product management system of the enterprise, or from a third-party data platform. Basic attributes include product name, brand, price, picture and description, etc. Sales statistics include sales quantity, sales amount, inventory and conversion rate, etc.

[0051] Specifically, the step of obtaining the product data of the product to be promoted can be implemented in the following manner: obtaining the product data of the product to be promoted by online data crawling, database query, or other corresponding methods such as direct user input. Further, based on the product data of the product to be promoted, a database about the product to be promoted is established. For example, the database only stores the information data of the product to be promoted, without considering the category to which it belongs. Of course, any other appropriate technical means can also be used to perform the corresponding processing work, which will not be repeated here.

[0052] S102: Determine at least two similar products of the product to be promoted, and determine product data of each of the at least two similar products.

[0053] The product to be promoted and the at least two similar products belong to the same product category.

[0054] Specifically, the method for determining at least one similar product may be as follows: first, a certain number of relevant web pages are obtained through a search engine or other web crawler technology, and then the contents of these web pages are analyzed and mined to extract some representative keywords, thereby determining the similarity of these keywords. If a certain keyword appears very frequently and its number of occurrences is much less than that of other keywords, then it can be considered that this one is a specific keyword. Therefore, by analyzing them, it is possible to find information sources that are closer to the true meaning. Of course, other various ways and methods can also be adopted to find relevant information and perform sorting and analysis, etc. to perform operations, and the above is only an example and does not limit the scope of the present invention. In addition, the identification of the same attributes and characteristics of the similar products, i.e., the products to be promoted, can also be achieved through a variety of different methods and technologies, so it will not be described in detail in this embodiment, and it is only necessary to ensure that they belong to the same category of products. Then, the weight coefficients of the corresponding similar products are calculated according to the corresponding product data of the at least two similar products determined above, and finally they are arranged in an array from high to low according to the size of the weight coefficient.

[0055] After determining the names of the two similar products that need to be compared, the first eigenvectors of the two similar products and the second eigenvector of the second similar product are constructed at the same time. After that, the Euclidean distance formula is used to compare the distance relationship between the above three eigenvectors to obtain the relative importance value of the multiple similar products (or the proportion of the entire group). It is worth noting that each individual here is not only an independent entity but also a component of many other elements. For example, it may be a person or organization unit of a company, institution or individual, or it may be just a team member or a single individual. Therefore, in actual application, it is necessary to take into account the different situations of the specific situation and then arrange and design it reasonably to ensure the maximum probability of the best effect.

[0056] S103: Calculate weight coefficients of at least two similar products, extract first k weight coefficients from all data weight coefficients by data sorting, and determine the similar products corresponding to the first k weight coefficients as recommended products.

[0057] k is an integer greater than or equal to 1.

[0058] Specifically, after the relative importance of the above-obtained similar products (that is, the numerical value of their proportion in the entire system) is arranged according to certain standards, they can be selected, and the following conditions must be met in this process: the numerical difference between any two non-adjacent similar products must not exceed the limit range of the set threshold. Since the average value is usually used to measure the role of each factor or indicator in its overall function, we assume here that the average value is selected as the reference point and the obtained number is doubled to make it just reach the upper limit requirement, and finally we get the ideal data we need. In addition, it should be noted that before this step, the size of the threshold to be used should be set first, and then it should be modified and adjusted according to actual needs to meet the standard of the expected goal.

[0059] In a possible implementation, the calculating the weight coefficients of the at least two similar products includes:

[0060] Constructing a first feature vector according to the product data of the product to be promoted;

[0061] constructing at least two second feature vectors according to the product data of the at least two similar products;

[0062] The distance between the first feature vector and each of the second feature vectors is calculated to obtain a weight coefficient of the at least two similar products.

[0063] First, construct the first feature vector (product to be promoted) and the second feature vector (similar products). Extract key features from the product data of the product to be promoted, such as brand, price, color, size, material, function, etc. Convert these feature values ​​into numerical form, for example, brand can be represented by code, price and size can be specific numerical values, color can be converted into numerical values ​​based on RGB values, etc. All numerical features are combined into a vector, namely the first feature vector. For each similar product, extract key features and convert them into numerical form to form a vector. In this way, multiple second feature vectors will be obtained, each corresponding to a similar product.

[0064] Next, choose a suitable distance metric, such as Euclidean distance (suitable for numerical features), cosine similarity (suitable for text features, such as product descriptions), or Manhattan distance (suitable for measuring the absolute difference between features). Calculate the distance between the feature vector of the product to be promoted and the feature vector of each similar product. The smaller the distance, the more similar the features of the two products are, and the higher the weight coefficient.

[0065] Then, the distance is converted into a weight coefficient. The following methods can be used to convert the distance into a weight coefficient: the reciprocal method (take the reciprocal of the distance, the smaller the distance, the larger the weight coefficient), the exponential decay method (use an exponential function to convert the distance into a weight coefficient, the smaller the distance, the larger the weight coefficient, and the faster the decay rate), or the linear mapping method (linearly map the distance to the interval [0,1], the smaller the distance, the larger the weight coefficient).

[0066] Finally, the same products are sorted according to the calculated weight coefficients, and the first k products with the highest weight coefficients are selected as recommended products from the sorted results.

[0067] S104: Displaying product data corresponding to the recommended product on the user's display desktop or browser page.

[0068] As an example, for mobile devices, you can develop native iOS or Android applications to display product data in a layout suitable for mobile devices. You can also develop mobile H5 pages based on web technology, using responsive design to ensure good display effects on screens of different sizes. Take advantage of the touch interaction characteristics of mobile devices to provide rich product display experiences such as picture carousels and left and right sliding.

[0069] As another example, you can use a cross-platform desktop application framework such as Electron or Qt to embed recommended product data into the application interface. Leverage local storage or caching technology to quickly load product data when the application starts, providing a smooth user experience. You can also implement lazy loading of product images to reduce initial loading time.

[0070] After completing the aforementioned steps S1 to S3, the next thing to do is to use conventional methods in the prior art (such as using a browser plug-in to automatically load) to present the prepared content in the agreed format on the user's host desktop or in the browser window for subsequent browsing and viewing.

[0071] The step of displaying the product data corresponding to the recommended product on the user's display desktop or browser page includes:

[0072] For each recommended product, the relevance score between the recommended product and the user is calculated based on the user's personal information and historical behavior data, and the recommended product with the highest score is displayed as the final recommended product on the user's display desktop or browser page.

[0073] First, collect user personal information, user preferences and historical behavior data, such as basic information such as age, gender, region, and hobbies, which can be obtained from user registration information or information actively filled in by users. User preferences can be analyzed and mined through data such as users' browsing history, search history, and purchase history on websites or apps, such as which brands, types of products, and products in which price ranges are frequently browsed by users. Historical behavior data may include user browsing history (recording which products users have browsed, the browsing time and dwell time of each product, etc.), user search history (recording which keywords users have searched, the search time and number of times, etc.), and user purchase history (recording which products users have purchased, the purchase time, quantity, and amount, etc.).

[0074] Then, calculate the relevance score between the recommended product and the user. You can use the following methods to calculate the relevance score between the recommended product and the user. You can choose or combine them according to the actual situation:

[0075] 1. Content similarity: The content similarity score is calculated by matching the features of the recommended products with the keywords in the user’s historical browsing and search history.

[0076] 2. Behavioral similarity: Based on the user’s historical browsing history, search history, and purchase history, analyze the user’s preference for similar products and calculate the behavioral similarity score.

[0077] 3. User portrait matching: Build a user portrait based on user personal information and historical behavior data, match recommended products with the user portrait, and calculate the matching score.

[0078] 4. Collaborative filtering: Based on the behavioral data of other users, such as users who have purchased the same product, recommend products that they may also be interested in and calculate the collaborative filtering score.

[0079] Finally, all recommended products are sorted according to their relevance scores, and the recommended product with the highest score is selected as the final recommended product.

[0080] Different recommendation strategies can be set according to different scenarios and goals. For example, when recommending on the home page, you can choose several products with the highest scores, and when recommending on the search page, you can choose several products with the highest relevance to the search keywords.

[0081] It is worth mentioning that although there are still some people who are unwilling to use such services because they are worried that certain illegal websites will threaten their privacy, with the development and progress of technology and the increase in security performance requirements, we have to make changes to adapt to the impact of this trend change.

[0082] However, it should be noted that although people often encounter various troubles in life, they cannot completely rely on this to solve all difficulties and obstacles, because in many cases, they will be restricted by the external environment, which will lead to the consequence that similar problems cannot be solved smoothly. On the other hand, after completing all the above preparations, you can officially enter the next stage of the workflow; that is, at this time, a new process should be started to replace the original running old application to achieve a seamless effect.

[0083] In this specific environment, there is usually a pre-installed operating system software package called Windows system, which is used to manage the internal hardware devices of the computer host. In addition, there are many optional options, including various development tools on the Linux / Unix platform and different types of projects such as server clusters. In short, as long as the basic development requirements are met, you can easily build a fully functional application framework in a short time to support the needs of future expansion functions; for the general public, you only need to be familiar with how to correctly use the existing resources to achieve the intended purpose without worrying too much about details and other information.

[0084] Obviously, in all the above scheme introductions, no specific implementation methods and related parameters of the algorithms are mentioned. This is because technicians in this technical field can make their own decisions and make appropriate improvements based on actual conditions to obtain better engineering application effects; and it does not rule out the possibility that there may be some potential values ​​that have not yet been discovered or fully recognized and need further research and exploration.

[0085] Therefore, the present invention generally provides a relatively novel and practical personalized advertising promotion solution. It not only simplifies the complexity of the original technology, but also reduces costs, improves efficiency, and increases security and other advantages in multiple dimensions. In addition, it is worth mentioning that no matter what the occasion, special attention should be paid to how to ensure the reliability and stability of the data transmission process in order to avoid damage or loss due to unexpected reasons. For this reason, it can be considered to use encryption technology and authentication mechanisms and other technical means to strengthen protection measures and minimize the possibility of risk loss.

[0086] In a possible implementation, the method further includes:

[0087] The user's response to the recommended product during the advertisement recommendation process is recorded, and the recommended product is adjusted according to the response result.

[0088] User response results include, for example:

[0089] Click: The user clicked on the recommended product link.

[0090] Viewed: The user viewed the detail page of the recommended product.

[0091] Add to Cart: User adds the recommended product to the shopping cart.

[0092] Purchase: The user purchased the recommended product.

[0093] Close Ad: The user closes or skips the recommended ad.

[0094] Negative feedback: A user expresses dissatisfaction with a recommended product, such as clicking a “Not Interested” button.

[0095] Recording methods: Various methods can be used to record user response results, such as: Website or APP logging system: record user visits, clicks on links, operation buttons, etc. User behavior analysis tools: real-time tracking and recording of user browsing, searching, purchasing, etc. User feedback mechanism: set up user feedback mechanisms, such as "dislike" buttons, comment functions, etc., to allow users to actively express their evaluation of recommended products.

[0096] After recording the response results, adjust the recommended products based on the response results

[0097] Adjustment strategy can adjust the recommendation strategy according to the user's response results, for example:

[0098] Increase weight: If a user has a positive response to a recommended product, such as clicking, browsing, or purchasing, the weight of the product can be increased, making it more likely to be recommended in the next recommendation.

[0099] Lower weight: If a user has a negative response to a recommended product, such as closing an ad or giving negative feedback, the weight of that product can be lowered, making it less likely to be recommended next time.

[0100] Add new products: You can discover other products that the user may be interested in based on their browsing history or purchase history and add them to the recommended product list.

[0101] Remove products: If a product has not been receiving a positive response from users, you can remove it from the list of recommended products.

[0102] The adjustment mechanism can adjust the recommended products using various methods, such as:

[0103] Machine learning algorithms: Use machine learning algorithms, such as reinforcement learning, to continuously adjust the parameters of the recommendation model based on user response results to improve recommendation results.

[0104] Rule Engine: Set some rules, for example, if a user clicks on a recommended product, this product will be given priority in the next recommendation.

[0105] A / B testing: Test different recommendation strategies, compare their effects, and select the best strategy.

[0106] The embodiment of the present application provides a personalized advertising recommendation method based on product data, obtains product data of products to be promoted, collects basic information and sales data of products to be promoted, and provides basic data for subsequent recommendation algorithms. By identifying similar products, more choices can be provided to users to avoid single recommendations. At the same time, collecting similar product data can help the algorithm analyze user needs more accurately. The weight of each similar product is calculated by analyzing similar product data, which reflects the relevance of each similar product to the user's interests. Sorting is performed according to the weight coefficient, and the similar products corresponding to the first k weight coefficients are selected as recommended products. This ensures the relevance and effectiveness of the recommended products. By displaying advertising information, users can easily obtain the latest recommendation information and improve the reach and user experience of advertisements.

[0107] The above are some specific implementations of the personalized advertising push method based on product data provided by the embodiment of the present application. Based on this, the present application also provides a corresponding device. The device provided by the embodiment of the present application will be introduced from the perspective of functional modularization.

[0108] See also Figure 2 The structure diagram of the personalized advertisement push device based on product data is shown, and the device includes an acquisition module 201, a determination module 202, a calculation module 203 and a display module 204.

[0109] The acquisition module 201 is used to acquire product data of the product to be promoted; the product data contains basic attributes of the product to be promoted and sales statistics of the product to be promoted;

[0110] The determination module 202 is used to determine at least two similar products of the product to be promoted, and to determine the product data of each of the at least two similar products; the product to be promoted and the at least two similar products belong to the same product category;

[0111] The calculation module 203 is used to calculate the weight coefficients of the at least two similar products, and extract the first k weight coefficients from all the data weight coefficients by data sorting, and determine the similar products corresponding to the first k weight coefficients as recommended products; k is an integer greater than or equal to 1;

[0112] The display module 204 is used to display the product data corresponding to the recommended product on the user's display desktop or browser page.

[0113] In a possible implementation, the calculation module 203 is specifically configured to:

[0114] Constructing a first feature vector according to the product data of the product to be promoted;

[0115] constructing at least two second feature vectors according to the product data of the at least two similar products;

[0116] The distance between the first feature vector and each of the second feature vectors is calculated to obtain a weight coefficient of the at least two similar products.

[0117] In a possible implementation, the display module 204 is specifically used for:

[0118] For each recommended product, the relevance score between the recommended product and the user is calculated based on the user's personal information and historical behavior data, and the recommended product with the highest score is displayed as the final recommended product on the user's display desktop or browser page.

[0119] In a possible implementation, the acquisition module 201 is specifically used for:

[0120] The product data of the product to be promoted is obtained by online data crawling, database query or user input.

[0121] In a possible implementation manner, the device further includes:

[0122] The adjustment module is used to record the user's response results to the recommended products during the advertisement recommendation process, and adjust the recommended products according to the response results.

[0123] The embodiments of the present application also provide corresponding devices and computer storage media for implementing the solutions provided by the embodiments of the present application.

[0124] The device includes a memory and a processor, the memory is used to store instructions or codes, and the processor is used to execute the instructions or codes so that the device executes the personalized advertising push method based on product data described in any embodiment of the present application.

[0125] The computer storage medium stores codes, and when the codes are executed, the device executing the codes implements the personalized advertisement push method based on product data as described in any embodiment of the present application.

[0126] The "first" and "second" (if any) in the names mentioned in the embodiments of the present application are only used as name identifiers and do not represent the first or second in order.

[0127] Through the description of the above implementation methods, it can be known that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a general hardware platform. Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a storage medium, such as a read-only memory (ROM) / RAM, a magnetic disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network communication device such as a router) to execute the methods described in each embodiment of the present application or some parts of the embodiments.

[0128] Each embodiment in this specification is described in a progressive manner, and the same and similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative work.

[0129] The above description is merely an exemplary embodiment of the present application and is not intended to limit the protection scope of the present application.

Claims

1. A personalized advertising push method based on product data, characterized in that: include: Acquire product data of the product to be promoted; the product data contains basic attributes of the product to be promoted and sales statistics of the product to be promoted; Determine at least two similar products of the product to be promoted, and determine product data of each of the at least two similar products; the product to be promoted and the at least two similar products belong to the same product category; Calculate the weight coefficients of the at least two similar products, extract the first k weight coefficients from all the data weight coefficients by data sorting, and determine the similar products corresponding to the first k weight coefficients as recommended products; k is an integer greater than or equal to 1; Displaying product data corresponding to the recommended product on the user's display desktop or browser page; The calculating the weight coefficients of the at least two similar products includes: Constructing a first feature vector according to the product data of the product to be promoted; constructing at least two second feature vectors according to the product data of the at least two similar products; Calculating the distance between the first feature vector and each of the second feature vectors to obtain a weight coefficient of the at least two similar products; The step of displaying the product data corresponding to the recommended product on the user's display desktop or browser page includes: For each recommended product, the relevance score between the recommended product and the user is calculated based on the user's personal information and historical behavior data, and the recommended product with the highest score is displayed as the final recommended product on the user's display desktop or browser page; The step of obtaining product data of the product to be promoted includes: Acquire product data of the product to be promoted by online data crawling, database query or user input; The method further comprises: The user's response to the recommended product during the advertisement recommendation process is recorded, and the recommended product is adjusted according to the response result.

2. A personalized advertising push device based on product data, characterized in that: include: An acquisition module, used to acquire product data of the product to be promoted; the product data contains basic attributes of the product to be promoted and sales statistics of the product to be promoted; A determination module, used to determine at least two similar products of the product to be promoted, and to determine the product data of each of the at least two similar products; the product to be promoted and the at least two similar products belong to the same product category; A calculation module, used to calculate the weight coefficients of the at least two similar products, and extract the first k weight coefficients from all the data weight coefficients by data sorting, and determine the similar products corresponding to the first k weight coefficients as recommended products; k is an integer greater than or equal to 1; A display module, used to display the product data corresponding to the recommended product on the user's display desktop or browser page; The calculation module is specifically used for: Constructing a first feature vector according to the product data of the product to be promoted; constructing at least two second feature vectors according to the product data of the at least two similar products; Calculating the distance between the first feature vector and each of the second feature vectors to obtain a weight coefficient of the at least two similar products; The display module is specifically used for: For each recommended product, the relevance score between the recommended product and the user is calculated based on the user's personal information and historical behavior data, and the recommended product with the highest score is displayed as the final recommended product on the user's display desktop or browser page; The acquisition module is specifically used for: Acquire product data of the product to be promoted by online data crawling, database query or user input; The device also includes: The adjustment module is used to record the user's response results to the recommended products during the advertisement recommendation process, and adjust the recommended products according to the response results.

3. A computer device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, personalized advertising push based on product data as described in claim 1 is implemented.

4. A computer storage medium, characterized in that: The computer storage medium stores instructions, and when the instructions are executed on the terminal device, the terminal device executes the personalized advertisement push based on product data as claimed in claim 1.

Images