A store ranking method and system based on private domain traffic
Through a store sorting method based on private domain traffic, combined with user registration and usage data, store ratings are calculated and user interests are combined, the problems of high costs and low efficiency in the existing technology are solved, and more accurate store sorting and user viscosity improvement are achieved.
Patent Information
- Application Number
- CN202411234370.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-09-04
AI Technical Summary
Existing e-commerce platforms have high cost and low efficiency in promotion and store sorting, especially for newly opened stores, it is difficult to accurately evaluate their quality, resulting in low user trust and increased resource consumption.
Using a store sorting method based on private domain traffic, by obtaining the data and usage data registered by users by scanning the store QR code, calculate the first and second scores of the store, and combine the user's interest types to comprehensively score and sort them, and set the target to top Stores to improve user viscosity and promotion efficiency.
It improves the accuracy of store sorting and user trust, reduces unnecessary resource consumption, and enhances platform promotion efficiency and user experience.
Smart Images

Figure CN118735614B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to a store ranking method and system based on private domain traffic. Background Art
[0002] With the booming development of e-commerce, consumers' shopping habits have gradually shifted from traditional retail to online shopping platforms. This shift has driven the fierce competition among platforms, and companies need to stand out in the fierce market competition through promotion activities. By promoting on the platform, companies can accurately target the target audience, increase brand exposure, and thus attract more potential consumers. Promotion activities also stimulate users' desire to buy and drive sales growth through promotions, discounts, and limited-time offers.
[0003] However, in the early stage of platform operation, the platform was promoted by ground promotion, which required hiring a large number of full-time ground promotion personnel in various communities and carrying out a series of preferential activities, such as giving specific small gifts or large coupons to new users after downloading the APP and registering as new users. However, for the early platforms, this high-cost promotion method is not conducive to the development of the platform. Most users uninstall the APP after registering new users and receiving corresponding rewards. The user stickiness is low, the promotion is not implemented, and the promotion efficiency is low.
[0004] On the other hand, for stores, the current ranking of stores in the platform is mainly based on user ratings of the stores. For some niche or newly opened stores with fewer visits from users, the store ratings are often difficult to truly determine the quality of the stores, resulting in users having lower trust in the ranking results and spending more time browsing stores, resulting in additional processing resource consumption on the platform and user devices. Summary of the invention
[0005] In order to solve the above-mentioned problems in the prior art, the present invention provides a store ranking method and system based on private domain traffic:
[0006] First aspect
[0007] The present invention provides a store ranking method based on private domain traffic, comprising:
[0008] S1: Acquire private domain traffic data, where the private domain traffic data includes: registration data and usage data of users completing platform registration by scanning a QR code displayed by a store;
[0009] S2: Determine the first score of each store based on the registration data of completing the platform registration by scanning the QR code displayed by the store;
[0010] S3: Determine the user's interest type based on the usage data;
[0011] S4: Determine a second score for each store based on the user's interest type;
[0012] S5: Determine a comprehensive score of each store according to the first score and the second score of each store;
[0013] S6: sorting the stores according to the comprehensive scores of the stores;
[0014] Among them, in the sorting results, the target store is placed at the top, and the target store is the store where the user completes the platform registration by scanning the QR code of the store.
[0015] Second aspect
[0016] The present invention provides a store ranking system based on private domain traffic, comprising:
[0017] An acquisition module, used to acquire private domain traffic data, wherein the private domain traffic data includes: registration data and usage data of a user completing platform registration by scanning a QR code displayed by a store;
[0018] A first rating module, for determining a first rating of each store based on registration data of a user who completes platform registration by scanning a QR code displayed by the store;
[0019] A determination module, configured to determine the user's interest type based on the usage data;
[0020] A second rating module, used to determine a second rating of each store based on the user's interest type;
[0021] A comprehensive scoring module, used to determine a comprehensive score of each store based on the first score and the second score of each store;
[0022] A sorting module, used to sort the stores according to the comprehensive scores of the stores;
[0023] Among them, in the sorting results, the target store is placed at the top, and the target store is the store where the user completes the platform registration by scanning the QR code of the store.
[0024] The beneficial effects of the present invention are reflected in that traffic is attracted by introducing users to download the platform client through stores, increasing user stickiness, and improving platform promotion efficiency. Stores that successfully introduce more users to download the platform client will also obtain higher first scores. Then, the stores are comprehensively ranked based on the second scores obtained in combination with the users' own interests, making the store ranking results more accurate, enhancing the users' trust in the ranking results, reducing unnecessary store traffic time, and reducing the additional processing resource consumption of the platform and user devices. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 A flow chart of a store ranking method based on private domain traffic provided by the present invention;
[0026] Figure 2 A schematic diagram of the structure of a store ranking method based on private domain traffic provided by the present invention;
[0027] Figure 3 A schematic diagram of a flow chart of a type determination method provided by the present invention;
[0028] Figure 4 A schematic flow chart of a method for calculating a second score provided by the present invention;
[0029] Figure 5 This is a schematic diagram of the structure of a store ranking system based on private domain traffic provided by the present invention. DETAILED DESCRIPTION
[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0031] Example 1
[0032] Reference Manual Attached Figure 1 , showing a flow chart of a store sorting method based on private domain traffic provided by the present invention.
[0033] Reference Manual Attached Figure 2 , showing a structural schematic diagram of a store ranking method based on private domain traffic provided by the present invention.
[0034] An embodiment of the present invention provides a store ranking method based on private domain traffic, comprising:
[0035] S1: Obtain private domain traffic data.
[0036] Among them, private domain traffic data refers to user data obtained by an enterprise or platform through its own channels and resources. Compared with public domain traffic (such as search engines, social media platforms, etc.), private domain traffic is more private and exclusive. Private domain traffic data includes various data generated by users in the channels of the enterprise or platform itself, mainly including registration data and usage data. The private domain data used in this invention is legal and compliant.
[0037] Optionally, private domain traffic data includes: registration data and usage data of users completing platform registration by scanning the QR code displayed in the store.
[0038] Alternatively, the enterprise or platform can display a QR code in the store to guide users to scan the code to download the APP or enter a specific page to complete the registration. The registration data may refer to the data of the user downloading and completing the platform registration through the QR code displayed in the store.
[0039] Optionally, usage data includes browsing history, purchase history, and search history. Information about products, stores, or pages that users have browsed on the platform, including browsing time, browsing duration, specific content browsed, etc., these data can reflect the user's interests and preferences. Records of users' purchase behavior on the platform, including purchased products, purchase time, purchase amount, etc. Purchase records are very important data that can be used to analyze users' consumption habits, popular products, etc. Users' search behavior on the platform, including search keywords, search time, search result clicks, etc. These data can help understand users' needs and concerns.
[0040] S2: Determine the first score of each store based on the registration data of the user who completes the platform registration by scanning the QR code displayed by the store.
[0041] In a possible implementation, S2 specifically includes: determining the first score of each store according to the following formula:
[0042]
[0043] in, Indicates i The first rating of the store, n i Indicates that the i The number of users who have completed platform registration using the QR code displayed in each store, N Indicates the total number of people who have completed platform registration within the preset time.
[0044] It should be noted that the score is calculated based on actual registration data, so it is more objective and accurate. The first score of each store directly reflects the number of users who completed the platform registration through the QR code displayed by the store within the preset time, so it is an objective measurement indicator.
[0045] Furthermore, the first score of a store is directly related to the number of users it successfully attracts. The more registered users a store successfully attracts, the higher the first score will be. In this way, the store will receive more incentives and rewards, which will increase the enthusiasm and motivation for attracting traffic to the store. At the same time, the first score can be a good evaluation of the activity and influence of the store. Registration data reflects the user's attention and interaction with the store, so a store with a high first score usually represents a high level of activity and popularity on the platform.
[0046] In the present invention, the benefits of determining the first score of a store based on registration data include objectivity, incentive effect, evaluation activity, guiding users to download the client, and data-driven decision-making. Such a scoring method helps the platform to better manage and operate stores, enhance the store's drainage effect and activity, and improve the platform's user experience and operational efficiency.
[0047] S3: Based on the usage data, determine the user's interest type.
[0048] Reference Manual Attached Figure 3 , which shows a flow chart of a type determination method provided by the present invention.
[0049] In a possible implementation, S3 specifically includes sub-steps S301 to S303:
[0050] S301: randomly shuffle the usage data and then structure it into multiple data samples.
[0051] S302: Input multiple data samples into a support vector machine.
[0052] Among them, Support Vector Machine (SVM) is a machine learning algorithm commonly used for classification and regression analysis, and is widely used in pattern recognition, data mining, predictive analysis and other fields. Its basic principle is to find an optimal hyperplane in high-dimensional space that can effectively separate data samples of different categories and maximize the interval between classification boundaries.
[0053] S303: Determine the judgment result of the interest type for each data sample based on the support vector machine.
[0054] It should be noted that the support vector machine is a binary classification method, which can adopt a cyclic judgment method to continuously detect whether each data sample belongs to a certain interest type, thereby completing the judgment of all interest types.
[0055] In the present invention, the benefits of judging the interest type of usage data based on support vector machines include personalized recommendations, accurate classification, processing nonlinear relationships, adapting to diverse data, and high efficiency. This method can provide users with more personalized and accurate recommendation services, improve the user experience of the platform and the success rate of product recommendations.
[0056] In a possible implementation manner, S303 specifically includes:
[0057] S3031: Based on the support vector machine, determine the judgment result of the interest type of each data sample through the following formula:
[0058]
[0059] in, x represents a data sample, represents the interest type judgment result obtained by the support vector machine model for the data sample x, w represents the weight, b represents the bias, Indicates that a dimension is The data is weighted w Make nonlinear transformations between lower and higher dimensional spaces and minimize the following objective function:
[0060]
[0061] in, represents the objective function, represents the support vector machine model, express The square of the modulus, represents the loss function, represents normalization, represents the jth data sample, Represents data samples According to the judgment results of Chat-gpt or other large language models on their interest types, Represents data samples The interest type judgment results obtained by the support vector machine model, M represents the total number of data samples.
[0062] It should be noted that through nonlinear transformation , SVM can handle nonlinear data. This method makes it possible to better find the optimal segmentation hyperplane in high-dimensional space for data of different interest types, thus improving the accuracy of classification.
[0063] Furthermore, the goal of SVM is to find an optimal hyperplane that maximizes the interval between the two types of data. The objective function in the formula uses Chat-gpt or other large language models to train samples. The judgment results of interest types replace the traditional manually labeled labels, which improves the generalization ability of the model and saves labor costs.
[0064] S3032: According to the judgment result of the interest type of each data sample, count the user's interest factor for each interest type:
[0065]
[0066] in, Indicates that the user k The interest factors of interest types, m k Indicates that the judgment result is k The number of data samples of interest for each interest type, M Represents the total number of data samples.
[0067] In the present invention, by counting the user's interest factors for each interest type, the user's interest preferences can be understood more accurately. This personalized interest analysis helps the platform recommend goods, services or content that are more in line with the user's preferences, improving the accuracy of the recommendation and user satisfaction. Understanding the user's interest factors for different interest types can help the platform optimize the user's browsing and experience. According to the user's interest preferences, the platform can adjust the page layout and content display, making it easier for users to find content that they are interested in, thereby improving the user's experience.
[0068] S4: Determine a second score for each store based on the user's interest type.
[0069] Reference Manual Attached Figure 4 , showing a flow chart of a method for calculating a second score provided by the present invention.
[0070] In a possible implementation, S4 specifically includes sub-steps S401 to S403:
[0071] S401: Obtain the business scope of each store.
[0072] S402: Determine the interest type of each store according to the business scope of each store.
[0073] S403: Based on the interest type to which each store belongs, the user's interest factor for the interest type is used as the second score of each store:
[0074]
[0075] in, Indicates i The second rating of the store, Indicatesi The interest type that the store belongs to, Indicates i The interest factor of the interest category to which the store belongs.
[0076] In the present invention, the user's interest factor for a certain interest type is used as the second score of the store, and personalized store sorting can be achieved. By showing the user the stores that they are interested in, the user's stay time and usage frequency can be increased. When users find that the platform provides stores that meet their interests, they will be more willing to browse and trade on the platform, which will enhance the stickiness and loyalty of users. Sorting stores according to the user's interest type can provide a better user experience. It will be easier for users to find stores and products that meet their needs, which will improve user satisfaction with the platform.
[0077] S5: Determine a comprehensive score of each store based on the first score and the second score of each store.
[0078] In a possible implementation, S5 specifically includes: determining the comprehensive score of each store according to the following formula:
[0079]
[0080] in, s i Indicates i The comprehensive score of each store, Indicates i The first rating of the store, Indicates i The second rating of the store.
[0081] In the present invention, the comprehensive score combines the first score and the second score, taking into full account the two factors of user registration volume and user interest matching. This can more comprehensively evaluate the pros and cons of each store and avoid the deviation caused by a single factor. Determining the comprehensive score of each store based on the first score and the second score of each store helps to comprehensively consider the impact of different factors on the store, optimize the recommendation ranking, improve the recommendation effect and user satisfaction, simplify the decision-making process, and enhance the credibility of the platform. This comprehensive scoring method can better help users find stores and products that meet their interests, promote the completion of transactions and the development of the platform.
[0082] In a possible implementation, S5 specifically includes: determining the comprehensive score of each store according to the following formula:
[0083]
[0084] in, s i Indicatesi The comprehensive score of each store, Indicates i The first rating of the store, Indicates i The second rating of the store, Represents the weight coefficient of the first score, and its value range is 0-1.
[0085] It should be noted that by introducing the weight coefficient of the first score , the importance of the first and second scores in the comprehensive score can be adjusted according to the actual situation. If the value is larger, more emphasis will be placed on the first score, otherwise, more emphasis will be placed on the second score. This flexibility allows the platform to make trade-offs based on specific circumstances and better meet actual needs.
[0086] In the present invention, the weight coefficient of the first score Flexible settings can balance different recommendation factors, optimize recommendation effects, enhance user experience, and simplify algorithms and decision-making processes. This comprehensive scoring method can better meet user needs and improve the platform's recommendation accuracy and user satisfaction.
[0087] S6: Sort the stores according to their comprehensive scores.
[0088] In the present invention, each store is sorted according to its comprehensive score, so that the store sorting result is more accurate, the user's trust in the sorting result is enhanced, unnecessary store traffic time is reduced, and the additional processing resource consumption of the platform and user equipment is reduced.
[0089] Among them, in the sorting results, the target store is placed at the top, and the target store is the store where the user completes the platform registration by scanning the QR code of the store.
[0090] It should be noted that users complete registration by scanning the QR code of the target store, which indicates that they have a certain interest in or recognition of the store. Placing the target store at the top meets the expectations and interests of users and improves user satisfaction and experience. On the other hand, placing the target store at the top can make it the first to be seen by users in the sorting results, greatly improving the exposure of the store, thereby encouraging the store to actively guide users to complete platform registration and improve the promotion efficiency of the platform.
[0091] In a possible implementation, the store ranking method based on private domain traffic further includes:
[0092] Rewards will be given to stores based on their first ratings.
[0093] In the present invention, issuing rewards to stores based on their first scores can motivate store competitiveness, increase store activity, improve user experience, promote transaction completion, and enhance the promotion effect of the platform. Such reward measures help the platform build a good ecosystem, making the stores on the platform more motivated and active, thereby promoting the sustainable development and growth of the platform.
[0094] The beneficial effects of the present invention are reflected in that traffic is attracted by introducing users to download the platform client through stores, increasing user stickiness, and improving platform promotion efficiency. Stores that successfully introduce more users to download the platform client will also obtain higher first scores. Then, the stores are comprehensively ranked based on the second scores obtained in combination with the users' own interests, making the store ranking results more accurate, enhancing the users' trust in the ranking results, reducing unnecessary store traffic time, and reducing the additional processing resource consumption of the platform and user devices.
[0095] Example 2
[0096] Reference Manual Attached Figure 5 , showing a structural schematic diagram of a store ranking system based on private domain traffic provided by the present invention.
[0097] An embodiment of the present invention provides a store ranking system 20 based on private domain traffic, comprising:
[0098] The acquisition module 201 is used to obtain private domain traffic data, which includes: registration data and usage data of users who complete platform registration by scanning the QR code displayed by the store.
[0099] The first scoring module 202 is used to determine the first score of each store based on the registration data of the user completing the platform registration by scanning the QR code displayed by the store.
[0100] The determination module 203 is used to determine the user's interest type based on the usage data.
[0101] The second scoring module 204 is used to determine a second score for each store based on the user's interest type.
[0102] The comprehensive scoring module 205 is used to determine the comprehensive score of each store according to the first score and the second score of each store.
[0103] A sorting module 206 is used to sort the stores according to their comprehensive scores;
[0104] Among them, in the sorting results, the target store is placed at the top, and the target store is the store where the user completes the platform registration by scanning the QR code of the store.
[0105] In a possible implementation, the first scoring module 202 is specifically configured to:
[0106] The first score of each store is determined according to the following formula:
[0107]
[0108] in, Indicates i The first rating of the store, n i Indicates that the i The number of users who have completed platform registration using the QR code displayed in each store, N Indicates the total number of people who have completed platform registration within the preset time.
[0109] In a possible implementation, the usage data includes browsing history, purchase history, and search history.
[0110] In a possible implementation, the determination module 203 is specifically configured to:
[0111] The usage data is randomly shuffled and then structured into multiple data samples.
[0112] Input multiple data samples into the support vector machine.
[0113] Based on the support vector machine, the judgment result of the interest type for each data sample is determined.
[0114] In a possible implementation, the determination module 203 is specifically configured to:
[0115] Based on the support vector machine, the judgment result of the interest type of each data sample is determined by the following formula:
[0116]
[0117] in, x represents a data sample, Represents data samples x The interest type judgment results obtained by the support vector machine model, w represents the weight, b represents the bias, Indicates that a dimension is The data is weighted w Make nonlinear transformations between lower and higher dimensional spaces and minimize the following objective function:
[0118]
[0119] in, represents the objective function, represents the support vector machine model, express The square of the modulus, represents the loss function, represents normalization, represents the jth data sample, Represents data samples According to the judgment results of Chat-gpt or other large language models on their interest types, Represents data samples The interest type judgment results obtained by the support vector machine model, M represents the total number of data samples.
[0120] According to the judgment results of the interest type of each data sample, the user's interest factor for each interest type is counted:
[0121]
[0122] in, Indicates that the user k The interest factors of interest types, m k Indicates that the judgment result is k The number of data samples of interest for each interest type, M Represents the total number of data samples.
[0123] In a possible implementation, the second scoring module 204 is specifically configured to:
[0124] Get the business scope of each store.
[0125] According to the business scope of each store, determine the interest type of each store.
[0126] Specifically, a large language model can be used to determine the interest type of each store based on the business scope of each store.
[0127] According to the interest type of each store, the user's interest factor for the interest type is used as the second score of each store:
[0128]
[0129] in, Indicates i The second rating of the store, Indicates i The interest type that the store belongs to, Indicates i The interest factor of the interest category to which the store belongs.
[0130] For example, if the store belongs to the snack category, the user's interest factor in the snack category will be used as the second score of the store.
[0131] In a possible implementation, the comprehensive scoring module 205 is specifically used for:
[0132] The comprehensive score of each store is determined according to the following formula:
[0133]
[0134] in, s i Indicates i The comprehensive score of each store, Indicates i The first rating of the store, Indicates i The second rating of the store.
[0135] In a possible implementation, the comprehensive scoring module 205 is specifically used for:
[0136] The comprehensive score of each store is determined according to the following formula:
[0137]
[0138] in, s i Indicates i The comprehensive score of each store, Indicates i The first rating of the store, Indicates i The second rating of the store, Represents the weight coefficient of the first score, and its value range is 0-1.
[0139] In a possible implementation, the store ranking system 20 based on private domain traffic further includes:
[0140] The distribution module is used to distribute rewards to stores based on the first ratings of each store.
[0141] A store ranking system 20 based on private domain traffic provided in an embodiment of the present invention can implement the steps and effects of the store ranking method based on private domain traffic in Example 1. To avoid repetition, the present invention will not go into details.
[0142] The beneficial effects of the present invention are reflected in that traffic is attracted by introducing users to download the platform client through stores, increasing user stickiness, and improving platform promotion efficiency. Stores that successfully introduce more users to download the platform client will also obtain higher first scores. Then, the stores are comprehensively ranked based on the second scores obtained in combination with the users' own interests, making the store ranking results more accurate, enhancing the users' trust in the ranking results, reducing unnecessary store traffic time, and reducing the additional processing resource consumption of the platform and user devices.
[0143] In the description of the embodiments of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "center", "top", "bottom", "top", "bottom", "inside", "outside", "inner side", "outer side" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. Among them, "inside" refers to an internal or enclosed area or space. "Periphery" refers to the area surrounding a specific component or a specific area.
[0144] In the description of the embodiments of the present invention, the terms "first", "second", "third", and "fourth" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", "third", and "fourth" may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.
[0145] In the description of the embodiments of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "install", "connect", "connect", and "assemble" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0146] In the description of the embodiments of the present invention, specific features, structures, materials or characteristics may be combined in a suitable manner in any one or more embodiments or examples.
[0147] In the description of the embodiments of the present invention, it should be understood that "-" and "~" represent the range between two values, and the range includes the endpoints. For example: "AB" represents a range greater than or equal to A and less than or equal to B. "A~B" represents a range greater than or equal to A and less than or equal to B.
[0148] In the description of the embodiments of the present invention, the term "and / or" herein is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " herein generally indicates that the associated objects before and after are in an "or" relationship.
[0149] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A store ranking method based on private domain traffic, characterized in that: include: S1: Acquire private domain traffic data, where the private domain traffic data includes: registration data and usage data of a user completing platform registration by scanning a QR code containing a store logo displayed by a store; S2: Determine the first score of each store based on the registration data of the user who completes the platform registration by scanning the QR code displayed by the store; S3: Determine the user's interest type based on the usage data; S4: Determine a second score for each store based on the user's interest type; S5: Determine a comprehensive score of each store according to the first score and the second score of each store; S6: sorting the stores according to the comprehensive scores of the stores; Among them, in the sorting results, the target store is placed at the top, and the target store is the store where the user completes the platform registration by scanning the QR code of the store; Wherein, the S2 is specifically: The first score of each store is determined according to the following formula: in, Indicates i The first rating of the store, n i Indicates that the i The number of users who have completed platform registration using the QR code displayed in each store, N Indicates the total number of people who have completed platform registration within the preset time period; Wherein, the S3 specifically includes: S301: randomly disrupting the usage data and then structuring it into a plurality of data samples; S302: Inputting a plurality of the data samples into a support vector machine; S303: Determine the judgment result of the interest type of each data sample based on the support vector machine; Wherein, the S303 specifically includes: S3031: Based on the support vector machine, determine the judgment result of the interest type of each data sample through the following formula: in, x represents a data sample, Represents data samples x The interest type judgment results obtained by the support vector machine model, w represents the weight, b Indicates bias, Indicates that a dimension is The data is weighted w Make nonlinear transformations between lower and higher dimensional spaces and minimize the following objective function: in, represents the objective function, represents the support vector machine model, express The square of the modulus, represents the loss function, represents normalization, Indicates j data samples, Represents data samples According to the judgment results of Chat-gpt or other large language models on their interest types, Represents data samples The interest type judgment results obtained by the support vector machine model, M Indicates the total number of data samples; S3032: According to the judgment result of the interest type of each data sample, count the user's interest factor for each interest type: in, Indicates that the user k The interest factors of interest types, m k Indicates that the judgment result is k The number of data samples of interest for each interest type, M Indicates the total number of data samples; Wherein, the S4 specifically includes: S401: Obtain the business scope of each store; S402: Determine the interest type to which each store belongs according to the business scope of each store; S403: Based on the interest type to which each store belongs, the user's interest factor for the interest type is used as the second score of each store: in, Indicates i The second rating of the store, Indicates i The interest type that the store belongs to, Indicates i The interest factor of the interest category to which the store belongs.
2. The store ranking method based on private domain traffic according to claim 1 is characterized in that: The usage data includes browsing history, purchase history and search history.
3. The store ranking method based on private domain traffic according to claim 1 is characterized in that: The S5 specifically includes: The comprehensive score of each store is determined according to the following formula: in, s i Indicates i The comprehensive score of each store, Indicates i The first rating of the store, Indicates i The second rating of the store.
4. The method for ranking stores based on private domain traffic according to claim 1, characterized in that: The S5 specifically includes: The comprehensive score of each store is determined according to the following formula: in, s i Indicates i The comprehensive score of each store, Indicates i The first rating of the store, Indicates i The second rating of the store, represents the weight coefficient of the first score, The value range is 0-1.
5. The method for ranking stores based on private domain traffic according to claim 1, characterized in that: Also includes: Rewards are issued to the stores based on the first scores of the stores.
6. A store ranking system based on private domain traffic, characterized in that: include: An acquisition module, used to acquire private domain traffic data, wherein the private domain traffic data includes: registration data and usage data of a user completing platform registration by scanning a QR code displayed by a store; A first rating module, for determining a first rating of each store based on registration data of a user who completes platform registration by scanning a QR code containing a store logo displayed by the store; A determination module, configured to determine the user's interest type based on the usage data; A second rating module, used to determine a second rating of each store based on the user's interest type; A comprehensive scoring module, used to determine a comprehensive score of each store based on the first score and the second score of each store; A sorting module, used to sort the stores according to the comprehensive scores of the stores; Among them, in the sorting results, the target store is placed at the top, and the target store is the store where the user completes the platform registration by scanning the QR code of the store; Wherein, the first scoring module is specifically used for: The first score of each store is determined according to the following formula: in, Indicates i The first rating of the store, n i Indicates that the i The number of users who have completed platform registration using the QR code displayed in each store, N Indicates the total number of people who have completed platform registration within the preset time period; Wherein, the determination module is specifically used for: The usage data is randomly disrupted and then structured into a plurality of data samples; Inputting a plurality of the data samples into a support vector machine; Based on the support vector machine, determine the judgment result of the interest type of each data sample; Wherein, the determination module is specifically used for: Based on the support vector machine, the judgment result of the interest type of each data sample is determined by the following formula: in, x represents a data sample, Represents data samples x The interest type judgment results obtained by the support vector machine model, w represents the weight, b Indicates bias, Indicates that a dimension is The data is weighted w Make nonlinear transformations between lower and higher dimensional spaces and minimize the following objective function: in, represents the objective function, represents the support vector machine model, express The square of the modulus, represents the loss function, represents normalization, Indicates j data samples, Represents data samples According to the judgment results of Chat-gpt or other large language models on their interest types, Represents data samples The interest type judgment results obtained by the support vector machine model, M Indicates the total number of data samples; According to the judgment results of the interest type of each data sample, the user's interest factor for each interest type is counted: in, Indicates that the user k The interest factors of interest types, m k Indicates that the judgment result is k The number of data samples of interest for each interest type, M Indicates the total number of data samples; Wherein, the second scoring module is specifically used for: Get the business scope of each store; Determine the interest type to which each store belongs based on its business scope; According to the interest type of each store, the user's interest factor for the interest type is used as the second score of each store: in, Indicates i The second rating of the store, Indicates i The interest type that the store belongs to, Indicates i The interest factor of the interest category to which the store belongs.
