A network interconnection method and system for batch advertising
By collecting basic advertising information for feature identification and cluster analysis, combining it with user consumption history databases for product demand and preference analysis, and utilizing an intelligent advertising user matching model to achieve precise matching between advertisements and users, the problem of insufficient accuracy in advertising network interconnection is solved, thereby improving the effectiveness of advertising.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- WEST WINDOW TECH (SUZHOU) CO LTD
- Filing Date
- 2022-11-07
- Publication Date
- 2026-07-28
AI Technical Summary
The accuracy of network interconnection in existing advertising technologies is insufficient, resulting in poor advertising performance.
By collecting basic advertising information, performing feature recognition and cluster analysis, a product advertising database is constructed. Combined with a user consumption history database, product demand and preference analysis is conducted, and an intelligent advertising user matching model is used to achieve accurate matching between advertisements and users.
It improves the accuracy and adaptability of network interconnection of advertisements, and enhances the precision and quality of advertisement delivery.
Smart Images
Figure CN116051198B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data processing, and more specifically, to a method and system for interconnecting networks for bulk advertising. Background Technology
[0002] Advertising, as a form of promotion, plays a vital role in corporate publicity, product sales, and information dissemination. In recent years, the quantity and types of advertisements have been continuously increasing, and the advertising market has expanded rapidly. However, the quality of advertisements varies greatly, and the effectiveness of advertising campaigns is often unsatisfactory. How to effectively place advertisements has become a major concern.
[0003] In existing technologies, there is a technical problem that the accuracy of network interconnection for advertising is insufficient, resulting in poor advertising performance. Summary of the Invention
[0004] This application provides a method and system for interconnecting networks for bulk advertising. It solves the technical problem of insufficient accuracy in network interconnection of advertisements in existing technologies, which leads to poor advertising delivery results. It achieves highly accurate and well-matched interconnection between advertisements and users, improving the accuracy and adaptability of network interconnection, enhancing the precision of advertising delivery, and ultimately improving the quality of advertising delivery.
[0005] In view of the above problems, this application provides a network interconnection method and system for bulk advertising.
[0006] In a first aspect, this application provides a method for interconnecting a network of bulk advertisements. The method is applied to a network interconnection system for bulk advertisements and includes: collecting basic information of multiple advertisements to obtain an advertisement database; performing feature recognition on the multiple advertisements based on the advertisement database to obtain multiple advertisement product feature recognition results, and performing cluster analysis on the advertisement database based on the multiple advertisement product feature recognition results to obtain a product advertisement database; performing advertisement quality evaluation on the multiple advertisements based on the product advertisement database to obtain multiple advertisement quality evaluation results; constructing a user consumption history database, performing product demand preference analysis on multiple users based on the user consumption history database to obtain multiple product demand preference analysis results; inputting the multiple advertisement product feature recognition results, the multiple advertisement quality evaluation results, and the multiple product demand preference analysis results into an intelligent advertisement user matching model to obtain multiple advertisement user matching results; and interconnecting the multiple advertisements and the multiple users through the advertisement network interconnection module based on the multiple advertisement user matching results and the product advertisement database.
[0007] Secondly, this application also provides a network interconnection system for bulk advertising, wherein the system includes: an advertising information collection module, which collects basic information of multiple advertisements to obtain an advertising database; an identification and analysis module, which performs feature identification on multiple advertisements based on the advertising database to obtain multiple advertising product feature identification results, and performs cluster analysis on the advertising database based on the multiple advertising product feature identification results to obtain a product advertising database; an advertising quality evaluation module, which performs advertising quality evaluation on the multiple advertisements based on the product advertising database to obtain multiple advertising quality evaluation results; and a product demand preference analysis module. The system comprises the following modules: a product demand preference analysis module, which constructs a user consumption history database and performs product demand preference analysis on multiple users based on this database to obtain multiple product demand preference analysis results; an advertising user matching module, which inputs the multiple advertising product feature identification results, the multiple advertising quality evaluation results, and the multiple product demand preference analysis results into an intelligent advertising user matching model to obtain multiple advertising user matching results; and an advertising user interconnection module, which interconnects the multiple advertising users and the multiple users through the advertising network interconnection module based on the multiple advertising user matching results and the product advertising database.
[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: By collecting basic information from multiple advertisements, an advertising database is obtained. Feature recognition of these advertisements yields product feature identification results, which are then clustered to create a product advertising database. Based on this database, advertising quality is evaluated, resulting in quality assessments. A user consumption history database is used to analyze user product demand preferences, generating product demand preference analysis results. These results are then input into an intelligent advertising user matching model to obtain matching results. Finally, based on the matching results and the product advertising database, an advertising network interconnection module connects the advertisements and users. This achieves highly accurate and well-matched interconnection between advertisements and users, improving the accuracy and adaptability of the advertising network, enhancing ad delivery precision, and ultimately improving overall ad delivery quality. Attached Figure Description
[0009] Figure 1 This is a flowchart illustrating a network interconnection method for bulk advertising according to this application; Figure 2 This is a schematic diagram illustrating the process of obtaining a product advertising database using a network interconnection method for bulk advertising, as described in this application. Figure 3 This is a flowchart illustrating the process of obtaining multiple product demand preference analysis results in a network interconnection method for bulk advertising according to this application; Figure 4 This is a schematic diagram of the structure of a network interconnection system for bulk advertising according to this application.
[0010] Explanation of reference numerals in the attached diagram: 11. Advertising information collection module; 12. Identification and analysis module; 13. Advertising quality assessment module; 14. Product demand and preference analysis module; 15. Advertising user matching module; 16. Advertising user interconnection module. Detailed Implementation
[0011] This application provides a method and system for interconnecting networks for bulk advertising. It solves the technical problem of insufficient accuracy in network interconnection of advertisements in existing technologies, which leads to poor advertising delivery results. It achieves highly accurate and well-matched interconnection between advertisements and users, improving the accuracy and adaptability of network interconnection, enhancing the precision of advertisement delivery, and ultimately improving the quality of advertisement delivery.
[0012] Example 1 Please see the appendix Figure 1 This application provides a method for interconnecting networks for bulk advertising, wherein the method is applied to a bulk advertising network interconnection system, the system including an advertising network interconnection module, and the method specifically includes the following steps: Step S100: Collect basic information from multiple advertisements to obtain an advertisement database; Specifically, basic information about multiple advertisements is collected through big data analysis to create an advertising database. This database includes basic information such as advertising slogans, content, videos, and themes for each advertisement. This achieves the technical goal of collecting basic information from multiple advertisements and creating an advertising database, laying the foundation for subsequent network interconnection of these advertisements.
[0013] Step S200: Based on the advertising database, perform feature recognition on multiple advertisements to obtain multiple advertising product feature recognition results, and perform cluster analysis on the advertising database based on the multiple advertising product feature recognition results to obtain a product advertising database; Further details are attached. Figure 2 As shown, step S200 of this application further includes: Step S210: Obtain multi-level advertising product feature dimensions, wherein the multi-level advertising product feature dimensions include advertising product type and advertising product price; Step S220: Based on the multi-level advertising product feature dimensions, perform feature recognition on the advertising database to obtain multiple advertising product feature recognition results, wherein the multiple advertising product feature recognition results include multiple advertising product type recognition results and multiple advertising product price recognition results; Step S230: Based on the identification results of the multiple advertising product types, perform cluster analysis on the advertising database to obtain the first-level cluster identification results; Step S240: Based on the price identification results of the multiple advertised products, perform cluster analysis on the first-level cluster identification results to obtain the product advertising database.
[0014] Specifically, feature identification is performed on the advertising database through multi-level advertising product feature dimensions to obtain multiple advertising product feature identification results, including multiple advertising product type identification results and multiple advertising product price identification results. Further, cluster analysis is performed on the advertising database according to the multiple advertising product type identification results to obtain first-level cluster identification results, and cluster analysis is performed on the first-level cluster identification results according to the multiple advertising product price identification results to obtain a product advertising database. The multi-level advertising product feature dimensions include advertising product type and advertising product price. Advertising product type refers to the product type information corresponding to the advertisement. Advertising product price refers to the product price information corresponding to the advertisement. The multiple advertising product type identification results include the advertising product type information corresponding to multiple advertisements in the advertising database. The multiple advertising product price identification results include the advertising product price information corresponding to multiple advertisements in the advertising database. The cluster analysis, when facing complex research objects, groups similar research objects into classes, maximizing the homogeneity or heterogeneity between classes, so that individuals within the same class have high similarity, and individuals in different classes have large differences. The first-level cluster identification results include multiple advertising product type clustering results. Each advertising product type cluster result includes multiple advertisements with the same advertising product type from the advertising database. The advertising product types differ significantly between different advertising product type cluster results. The product advertising database includes multiple advertising product type-price cluster results. Multiple advertising product type-price cluster results correspond to multiple advertising product type cluster results with the same advertising product price. That is, each advertising product type-price cluster result includes multiple advertisements in the advertising database that have the same advertising product type and advertising product price. This achieves the technical effect of constructing a product advertising database by performing feature identification and cluster analysis on the advertising database, thereby improving the accuracy and adaptability of interconnecting multiple advertisements on the network.
[0015] Step S300: Based on the product advertising database, evaluate the advertising quality of the multiple advertisements to obtain multiple advertising quality evaluation results; Furthermore, step S300 of this application also includes: Step S310: Based on the product advertising database, construct an advertising quality evaluation database, wherein the advertising quality evaluation database includes a dataset of historical sales records of advertised products and a dataset of historical browsing records of advertisements; Specifically, based on a product advertising database, historical information is collected from multiple advertisements in the database to obtain a dataset of historical sales records for advertised products and a dataset of historical browsing records for advertisements, thereby establishing an advertising quality assessment database. The advertising quality assessment database includes the dataset of historical sales records for advertised products and the dataset of historical browsing records for advertisements. The dataset of historical sales records for advertised products includes sales data such as historical sales volume for multiple products corresponding to multiple advertisements. The dataset of historical browsing records for advertisements includes browsing data such as historical number of views, historical browsing duration, historical number of forwards, and historical number of comments for multiple advertisements. This achieves the technical effect of constructing an advertising quality assessment database, providing reliable data references for subsequent advertising quality assessments of multiple advertisements.
[0016] Step S320: Based on the historical sales record dataset of the advertised products, evaluate the sales influence of the multiple advertisements to obtain multiple advertisement sales influence coefficients; Furthermore, step S320 of this application also includes: Step S321: Based on the historical sales record dataset of the advertised products and the multiple advertisements, perform sales correlation analysis to obtain multiple advertisement sales correlation coefficients; Step S322: Obtain a preset advertising sales correlation coefficient, and filter the plurality of advertising sales correlation coefficients based on the preset advertising sales correlation coefficient to obtain a plurality of high-quality advertising sales correlation coefficients that are not less than the preset advertising sales correlation coefficient; Step S323: Match the historical sales record dataset of the advertised products based on the multiple high-quality advertising sales correlation coefficients to obtain an advertising-related sales record dataset; Step S324: Based on the advertising-related sales record dataset, evaluate the sales influence of the multiple advertisements to obtain the sales influence coefficients of the multiple advertisements.
[0017] Specifically, a sales correlation analysis is performed on the historical sales record dataset of the advertised products and multiple advertisements to obtain multiple advertising sales correlation coefficients. These coefficients are then filtered according to a preset set of correlation coefficients to obtain multiple high-quality correlation coefficients. Further, the historical sales record dataset of the advertised products is matched against these high-quality correlation coefficients to obtain an advertising-related sales record dataset. Subsequently, the sales influence of multiple advertisements is evaluated based on this dataset, yielding multiple advertising sales influence coefficients. These multiple advertising sales correlation coefficients are parameters characterizing the correlation between the historical sales volume of multiple products and multiple advertisements within the historical sales record dataset. The greater the correlation between a product's historical sales volume and an advertisement, the larger the corresponding advertising sales correlation coefficient. The preset advertising sales correlation coefficient is a pre-defined threshold. The multiple high-quality correlation coefficients include multiple advertising sales correlation coefficients that are equal to or greater than the preset correlation coefficient. The advertising-related sales record dataset includes the sales record data corresponding to the multiple high-quality correlation coefficients within the historical sales record dataset of the advertised products. The multiple advertising sales influence coefficients are parameter information used to characterize the sales influence of multiple advertisements. For example, the more historical sales volume of a product in the advertising-related sales record dataset, the more products are sold under the influence of the corresponding advertisement, the greater the sales influence of the advertisement, and the higher the corresponding advertising sales influence coefficient. This achieves the technical effect of evaluating the sales influence of multiple advertisements using a dataset of historical sales records of advertised products, obtaining multiple advertising sales influence coefficients, and improving the accuracy of advertising quality evaluation for multiple advertisements.
[0018] Step S330: Based on the advertising history browsing record dataset, evaluate the popularity of the multiple advertisements to obtain multiple advertising popularity evaluation coefficients; Furthermore, step S330 of this application also includes: Step S331: Obtain a preset set of advertising popularity evaluation features, wherein the preset set of advertising popularity evaluation features includes multiple preset advertising popularity evaluation features and multiple preset advertising popularity evaluation feature values; Step S332: Based on the preset set of advertising popularity evaluation features, perform feature recognition on the historical browsing record dataset of advertisements to obtain the advertising popularity feature recognition result; Step S333: Based on the advertising popularity feature recognition results, obtain the multiple advertising popularity evaluation coefficients.
[0019] Step S340: Obtain the preset weight allocation conditions; Step S350: Based on the preset weight allocation conditions, perform weighted calculations on the multiple advertising sales influence coefficients and the multiple advertising popularity evaluation coefficients to obtain the multiple advertising quality evaluation results.
[0020] Specifically, the historical browsing record dataset of advertisements is subjected to feature identification according to a preset set of advertisement popularity evaluation features to obtain advertisement popularity feature identification results, and then multiple advertisement popularity evaluation coefficients are determined. Further, according to preset weight allocation conditions, multiple advertisement sales influence coefficients and multiple advertisement popularity evaluation coefficients are weighted and calculated to obtain multiple advertisement quality evaluation results. The preset set of advertisement popularity evaluation features includes multiple preset advertisement popularity evaluation features and multiple preset advertisement popularity evaluation feature values. The multiple preset advertisement popularity evaluation features include pre-set information on multiple advertisement popularity evaluation features. The multiple preset advertisement popularity evaluation feature values include preset feature value information corresponding to the multiple preset advertisement popularity evaluation features. For example, the multiple preset advertisement popularity evaluation features include an advertisement historical browsing duration of 30 seconds and an advertisement historical browsing count of 100 times. The multiple preset advertisement popularity evaluation feature values include a preset feature value 1 corresponding to an advertisement historical browsing duration of 30 seconds and a preset feature value 10 corresponding to an advertisement historical browsing count of 100 times. The advertisement popularity feature identification results include multiple sets of advertisement popularity feature identification. The multiple ad popularity feature recognition sets include preset ad popularity evaluation features and preset ad popularity evaluation feature values corresponding to the browsing history data of multiple ads in the ad historical browsing record dataset. The multiple ad popularity evaluation coefficients include the sum of the preset ad popularity evaluation feature values corresponding to the multiple ad popularity feature recognition sets. The preset weight allocation conditions include pre-set and determined sales influence weight allocation coefficients and ad popularity weight allocation coefficients. The multiple ad quality evaluation results include multiple weighted calculation results obtained by weighting the multiple ad sales influence coefficients and multiple ad popularity evaluation coefficients according to the preset weight allocation conditions. For example, the sales influence weight allocation coefficient is α, and the ad popularity weight allocation coefficient is β. Among the multiple ads, the ad sales influence coefficient corresponding to ad A is... The popularity evaluation coefficient for advertisement A is: Multiple ad quality assessment results include the weighted calculation result corresponding to ad A. Weighted calculation results for +β This technology achieves the goal of improving the accuracy and adaptability of network interconnection of advertisements by weighting multiple advertisement sales influence coefficients and multiple advertisement popularity evaluation coefficients through preset weight allocation conditions, thereby obtaining reliable multiple advertisement quality evaluation results.
[0021] Step S400: Construct a user consumption history database, and perform product demand preference analysis on multiple users based on the user consumption history database to obtain multiple product demand preference analysis results; Further details are attached. Figure 3 As shown, step S400 of this application further includes: Step S410: Obtain a preset set of historical times; Step S420: Based on the preset historical time set, collect consumption history information from the multiple users to obtain the user consumption history database; Step S430: Based on the user consumption history database, obtain information on multiple consumer product types; Step S440: Based on the multiple consumer product type information, perform consumption distribution statistics on the user consumption history database to obtain multiple consumption distribution statistics results; Step S450: Based on the multiple consumption distribution statistics, evaluate the consumption popularity of the multiple consumer product types to obtain multiple consumption popularity evaluation results; Step S460: Based on the multiple consumer popularity assessment results, mark the multiple consumer product type information to obtain the multiple product demand preference analysis results.
[0022] Specifically, the consumption history information of multiple users is collected according to a preset historical time set to obtain a user consumption history database. Further, multiple consumption product type information is extracted from the user consumption history database, and consumption amount and frequency are statistically analyzed according to these multiple consumption product type information to obtain multiple consumption distribution statistical results. Then, based on the multiple consumption distribution statistical results, the consumption popularity of the multiple consumption product type information is evaluated to obtain multiple consumption popularity evaluation results. These results are then used to label the multiple consumption product type information to obtain multiple product demand preference analysis results. The preset historical time set includes multiple pre-set historical time nodes. For example, the preset historical time set includes historical time nodes such as 3 days ago, 1 week ago, and 3 months ago. The user consumption history database includes multiple sets of consumption history information corresponding to multiple users under the preset historical time set. Each set of consumption history information includes multiple consumption history information for each user under the preset historical time set. The multiple consumption product type information includes multiple sets of consumption product types. Each set of consumption product type information in the user consumption history database corresponds to a specific consumption product type. The multiple consumption distribution statistics include multiple consumption distribution statistical sets. These sets include consumption amount and frequency information corresponding to multiple product type sets in the user's consumption history database. The multiple consumption popularity assessment results include multiple consumption distribution statistics and consumption popularity parameters corresponding to multiple product type information. Higher consumption amounts and frequency result in higher consumption popularity parameters for the corresponding product type information. The multiple product demand preference analysis results include multiple consumption popularity assessment results, multiple product type information, and the correspondence between these results. This achieves the technical effect of improving the accuracy of network interconnection of advertising by analyzing product demand preferences of multiple users through the user's consumption history database and obtaining multiple product demand preference analysis results.
[0023] Step S500: Input the multiple advertising product feature identification results, the multiple advertising quality assessment results, and the multiple product demand preference analysis results into the intelligent advertising user matching model to obtain multiple advertising user matching results; Furthermore, step S500 of this application also includes: Step S510: The intelligent advertising user matching model includes an input layer, an advertising incentive adjustment layer, an advertising user matching layer, and an output layer; Step S520: Input the multiple advertising quality evaluation results into the advertising incentive adjustment layer to obtain multiple advertising incentive adjustment coefficients; Step S530: Input the multiple advertising incentive adjustment coefficients, the multiple advertising product feature identification results, and the multiple product demand preference analysis results into the advertising user matching layer to obtain the multiple advertising user matching results.
[0024] Step S600: Based on the multiple advertising user matching results and the product advertising database, the multiple advertisements and the multiple users are interconnected through the advertising network interconnection module.
[0025] Specifically, multiple advertising quality assessment results are used as input information and fed into the advertising incentive adjustment layer of the intelligent advertising user matching model to obtain multiple advertising incentive adjustment coefficients. Then, multiple advertising incentive adjustment coefficients, multiple advertising product feature identification results, and multiple product demand preference analysis results are used as input information and fed into the advertising user matching layer of the intelligent advertising user matching model to obtain multiple advertising user matching results. These results are then output through the output layer. Combined with a product advertising database, the multiple advertisements and multiple users are interconnected using an advertising network interconnection module. The intelligent advertising user matching model includes an input layer, an advertising incentive adjustment layer, an advertising user matching layer, and an output layer. The advertising incentive adjustment layer is trained using a large amount of data related to the advertising quality assessment results and has the function of analyzing and matching incentive adjustment coefficients to the multiple input advertising quality assessment results. The higher the advertising quality assessment result, the stronger the sales influence and the higher the popularity of the corresponding advertisement, and the higher the incentive adjustment coefficient. The advertising product feature identification result corresponding to that advertising quality assessment result is used preferentially for user matching. The multiple advertising incentive adjustment coefficients are parameters used to characterize the matching order of multiple advertising product feature identification results corresponding to multiple advertising quality assessment results. The higher the incentive adjustment coefficient, the earlier the matching order of the advertising product feature identification result corresponding to that incentive adjustment coefficient. The advertising user matching layer is trained with a large amount of data related to advertising incentive adjustment coefficients, advertising product feature identification results, and product demand preference analysis results, and has the function of intelligently analyzing and matching multiple input advertising product feature identification results and multiple product demand preference analysis results. The multiple advertising user matching results include the correspondence between multiple product demand preference analysis results and multiple advertising product feature identification results under multiple advertising incentive adjustment coefficients. For example, product demand preference analysis result B corresponds to advertising product feature identification results b1 and b2. However, the advertising quality assessment result corresponding to advertising product feature identification result b1 is higher, and the incentive adjustment coefficient is larger. Therefore, the obtained multiple advertising user matching results include advertising product feature identification results b2 and b1 corresponding to product demand preference analysis result B. The advertising network interconnection module, included in the batch advertising network interconnection system, has the function of interconnecting multiple users and multiple advertisements in the product advertising database according to multiple advertising user matching results. That is, the advertising network interconnection module can push multiple advertisements in the product advertising database to multiple users according to multiple advertising user matching results.The technology achieves the goal of accurately and efficiently matching and analyzing multiple advertising product feature identification results, multiple advertising quality assessment results, and multiple product demand preference analysis results through an intelligent advertising user matching model, obtaining multiple advertising user matching results, and interconnecting multiple advertisements and multiple users according to the multiple advertising user matching results, thereby improving the accuracy of advertising network interconnection and thus improving the quality of advertising delivery.
[0026] In summary, the network interconnection method for bulk advertising provided in this application has the following technical effects: 1. By collecting basic information from multiple advertisements, an advertising database is obtained. Feature recognition of these advertisements yields product feature identification results, which are then clustered to create a product advertising database. Based on this database, advertising quality is evaluated, resulting in quality assessments. A user consumption history database is used to analyze user product needs and preferences, generating product demand and preference analysis results. These results are then input into an intelligent advertising user matching model to obtain matching results. Finally, based on the matching results and the product advertising database, an advertising network interconnection module connects the advertisements and users. This achieves highly accurate and well-matched interconnection between advertisements and users, improving the accuracy and adaptability of the advertising network, enhancing ad delivery precision, and ultimately improving overall ad delivery quality.
[0027] 2. By using preset weighting conditions to perform weighted calculations on multiple advertising sales influence coefficients and multiple advertising popularity evaluation coefficients, reliable multiple advertising quality evaluation results are obtained, thereby improving the accuracy and adaptability of advertising network interconnection.
[0028] 3. By analyzing the product demand preferences of multiple users through a user consumption history database, the analysis results of multiple product demand preferences can be obtained, thereby improving the accuracy of network interconnection of advertisements.
[0029] 4. Through the intelligent advertising user matching model, the system accurately and efficiently matches and analyzes multiple advertising product feature identification results, multiple advertising quality assessment results, and multiple product demand preference analysis results to obtain multiple advertising user matching results. Based on these multiple advertising user matching results, the system interconnects multiple advertisements and multiple users, thereby improving the accuracy of advertising network interconnection and thus enhancing the quality of advertising delivery.
[0030] Example 2 Based on the same inventive concept as the network interconnection method for bulk advertising in the foregoing embodiments, this invention also provides a network interconnection system for bulk advertising. Please refer to the appendix. Figure 4 The system includes: Advertising information collection module 11 is used to collect basic information of multiple advertisements to obtain an advertising database; The identification and analysis module 12 is used to perform feature identification on multiple advertisements based on the advertising database, obtain multiple advertising product feature identification results, and perform cluster analysis on the advertising database based on the multiple advertising product feature identification results to obtain a product advertising database. The advertising quality assessment module 13 is used to assess the advertising quality of the multiple advertisements based on the product advertising database and obtain multiple advertising quality assessment results. Product demand preference analysis module 14 is used to construct a user consumption history database, perform product demand preference analysis on multiple users based on the user consumption history database, and obtain multiple product demand preference analysis results; The advertising user matching module 15 is used to input the multiple advertising product feature identification results, the multiple advertising quality evaluation results, and the multiple product demand preference analysis results into the intelligent advertising user matching model to obtain multiple advertising user matching results. The advertising user interconnection module 16 is used to interconnect the multiple advertisements and the multiple users through the advertising network interconnection module based on the multiple advertising user matching results and the product advertising database.
[0031] Furthermore, the system also includes: The feature dimension acquisition module is used to obtain multi-level advertising product feature dimensions, wherein the multi-level advertising product feature dimensions include advertising product type and advertising product price; An advertising feature recognition module is used to perform feature recognition on the advertising database based on the multi-level advertising product feature dimensions to obtain multiple advertising product feature recognition results, wherein the multiple advertising product feature recognition results include multiple advertising product type recognition results and multiple advertising product price recognition results; A first-level clustering identification result acquisition module is used to perform clustering analysis on the advertising database based on the identification results of the multiple advertising product types to obtain first-level clustering identification results; The product advertising database acquisition module is used to perform cluster analysis on the first-level clustering identification results based on the multiple advertising product price identification results to obtain the product advertising database.
[0032] Furthermore, the system also includes: An advertising quality assessment database construction module is used to construct an advertising quality assessment database based on the product advertising database. The advertising quality assessment database includes a dataset of historical sales records of advertised products and a dataset of historical browsing records of advertisements. The sales influence assessment module is used to assess the sales influence of the multiple advertisements based on the historical sales record dataset of the advertised products, and obtain the sales influence coefficients of the multiple advertisements. A welcome evaluation module is used to evaluate the welcome of the multiple advertisements based on the advertisement history browsing record dataset, and obtain multiple advertisement welcome evaluation coefficients; A condition acquisition module, which is used to acquire preset weight allocation conditions; The advertising quality assessment result acquisition module is used to perform weighted calculations on the multiple advertising sales influence coefficients and the multiple advertising popularity assessment coefficients based on the preset weight allocation conditions, and obtain the multiple advertising quality assessment results.
[0033] Furthermore, the system also includes: A sales correlation analysis module is used to perform sales correlation analysis based on the historical sales record dataset of the advertised products and the multiple advertisements to obtain multiple advertisement sales correlation coefficients. A sales correlation coefficient filtering module is used to obtain a preset advertising sales correlation coefficient, and to filter the plurality of advertising sales correlation coefficients based on the preset advertising sales correlation coefficient to obtain a plurality of high-quality advertising sales correlation coefficients that are not less than the preset advertising sales correlation coefficient. The dataset matching module is used to match the historical sales record dataset of the advertising product based on the multiple high-quality advertising sales correlation coefficients to obtain the advertising-related sales record dataset. The advertising sales influence coefficient acquisition module is used to evaluate the sales influence of the multiple advertisements based on the advertising-related sales record dataset and obtain the sales influence coefficients of the multiple advertisements.
[0034] Furthermore, the system also includes: An evaluation feature set acquisition module is used to obtain a preset advertising popularity evaluation feature set, wherein the preset advertising popularity evaluation feature set includes multiple preset advertising popularity evaluation features and multiple preset advertising popularity evaluation feature values; The module for obtaining advertising popularity feature recognition results is used to perform feature recognition on the historical browsing record dataset of advertisements based on the preset advertising popularity evaluation feature set, and obtain advertising popularity feature recognition results. An advertising popularity evaluation coefficient acquisition module is used to obtain the plurality of advertising popularity evaluation coefficients based on the advertising popularity feature recognition results.
[0035] Furthermore, the system also includes: A historical time determination module is used to obtain a preset set of historical times. The user consumption history database acquisition module is used to collect consumption history information of the multiple users based on the preset historical time set to obtain the user consumption history database. A consumer product type determination module is used to obtain information on multiple consumer product types based on the user's consumption history database. The consumption distribution statistics module is used to perform consumption distribution statistics on the user consumption history database based on the multiple consumer product type information, and obtain multiple consumption distribution statistics results; A consumer popularity assessment module is used to assess the consumer popularity of multiple consumer product types based on the multiple consumer distribution statistics, and obtain multiple consumer popularity assessment results. The product demand preference analysis result acquisition module is used to mark the multiple consumer product type information based on the multiple consumer popularity assessment results, and obtain the multiple product demand preference analysis results.
[0036] Furthermore, the system also includes: The model component module, used in the intelligent advertising user matching model, includes an input layer, an advertising incentive adjustment layer, an advertising user matching layer, and an output layer; An advertising incentive adjustment coefficient acquisition module is used to input the multiple advertising quality evaluation results into the advertising incentive adjustment layer to obtain multiple advertising incentive adjustment coefficients. The advertising user matching result acquisition module is used to input the multiple advertising incentive adjustment coefficients, the multiple advertising product feature identification results, and the multiple product demand preference analysis results into the advertising user matching layer to obtain the multiple advertising user matching results.
[0037] This application provides a method for interconnecting a network of bulk advertisements. The method is applied to a network interconnection system for bulk advertisements. The method includes: collecting basic information from multiple advertisements to obtain an advertisement database; performing feature recognition on the multiple advertisements to obtain multiple product feature recognition results, and performing cluster analysis on the advertisement database according to the multiple product feature recognition results to obtain a product advertisement database; evaluating the quality of the multiple advertisements based on the product advertisement database to obtain multiple advertisement quality evaluation results; analyzing the product demand preferences of multiple users through a user consumption history database to obtain multiple product demand preference analysis results; inputting the multiple advertisement product feature recognition results, multiple advertisement quality evaluation results, and multiple product demand preference analysis results into an intelligent advertisement user matching model to obtain multiple advertisement user matching results; and interconnecting the multiple advertisements and multiple users through an advertisement network interconnection module based on the multiple advertisement user matching results and the product advertisement database. This method solves the technical problem of insufficient accuracy in network interconnection of advertisements in existing technologies, which leads to poor advertising delivery results. It achieves high-precision and highly matched interconnection between advertisements and users, improving the accuracy and adaptability of network interconnection of advertisements, enhancing the precision of advertisement delivery, and thus improving the quality of advertisement delivery.
[0038] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0039] This specification and accompanying drawings are merely illustrative examples of this application. If any modifications and variations of this invention fall within the scope of this invention and its equivalents, this invention also intends to include such modifications and variations.
Claims
1. A network interconnection method for batch advertising, characterized by, The method is applied to a bulk advertising network interconnection system, the system including an advertising network interconnection module, and the method includes: Collect basic information from multiple advertisements to obtain an advertisement database; Based on the advertising database, feature identification is performed on multiple advertisements to obtain multiple advertising product feature identification results. Then, based on the multiple advertising product feature identification results, cluster analysis is performed on the advertising database to obtain a product advertising database. Based on the product advertising database, the advertising quality of the multiple advertisements is evaluated to obtain multiple advertising quality evaluation results; Construct a user consumption history database, and perform product demand preference analysis on multiple users based on the user consumption history database to obtain multiple product demand preference analysis results; The multiple advertising product feature identification results, the multiple advertising quality assessment results, and the multiple product demand preference analysis results are input into the intelligent advertising user matching model to obtain multiple advertising user matching results; Based on the multiple advertising user matching results and the product advertising database, the multiple advertisements and the multiple users are interconnected through the advertising network interconnection module; The method further includes: evaluating the quality of the multiple advertisements based on the product advertising database to obtain multiple advertising quality evaluation results; and the method also includes: Based on the product advertising database, an advertising quality evaluation database is constructed, which includes a dataset of historical sales records of advertised products and a dataset of historical browsing records of advertisements. Based on the historical sales record dataset of the advertised products, the sales influence of the multiple advertisements is evaluated to obtain multiple advertisement sales influence coefficients. Based on the aforementioned historical browsing record dataset, the popularity of the multiple advertisements is evaluated to obtain multiple advertisement popularity evaluation coefficients; Obtain the preset weight allocation conditions; Based on the preset weight allocation conditions, the multiple advertising sales influence coefficients and multiple advertising popularity evaluation coefficients are weighted and calculated to obtain the multiple advertising quality evaluation results. The method further includes, based on the historical sales record dataset of the advertised products, evaluating the sales influence of the multiple advertisements to obtain multiple advertisement sales influence coefficients; Based on the historical sales record dataset of the advertised products and the multiple advertisements, a sales correlation analysis is performed to obtain multiple advertisement sales correlation coefficients; Obtain a preset advertising sales correlation coefficient, and based on the preset advertising sales correlation coefficient, filter the plurality of advertising sales correlation coefficients to obtain a plurality of high-quality advertising sales correlation coefficients that are not less than the preset advertising sales correlation coefficient; The historical sales record dataset of the advertised products is matched with the multiple high-quality advertising sales correlation coefficients to obtain an advertising-related sales record dataset. Based on the aforementioned sales record dataset associated with the advertisements, the sales influence of the multiple advertisements is evaluated to obtain the sales influence coefficients of the multiple advertisements. The method further includes, based on the historical browsing record dataset of the advertisements, evaluating the popularity of the multiple advertisements to obtain multiple advertisement popularity evaluation coefficients. Obtain a preset set of advertising popularity evaluation features, wherein the preset set of advertising popularity evaluation features includes multiple preset advertising popularity evaluation features and multiple preset advertising popularity evaluation feature values; Based on the preset set of advertising popularity evaluation features, feature recognition is performed on the historical browsing record dataset of advertisements to obtain the advertising popularity feature recognition result; Based on the advertising popularity feature recognition results, the multiple advertising popularity evaluation coefficients are obtained; The method for obtaining multiple advertising user matching results further includes: The intelligent advertising user matching model includes an input layer, an advertising incentive adjustment layer, an advertising user matching layer, and an output layer; The multiple advertising quality evaluation results are input into the advertising incentive adjustment layer to obtain multiple advertising incentive adjustment coefficients; The multiple advertising incentive adjustment coefficients, the multiple advertising product feature identification results, and the multiple product demand preference analysis results are input into the advertising user matching layer to obtain the multiple advertising user matching results.
2. The method of claim 1, wherein, The method for obtaining the product advertising database further includes: Obtain multi-level advertising product feature dimensions, wherein the multi-level advertising product feature dimensions include advertising product type and advertising product price; Based on the multi-level advertising product feature dimensions, feature recognition is performed on the advertising database to obtain multiple advertising product feature recognition results, wherein the multiple advertising product feature recognition results include multiple advertising product type recognition results and multiple advertising product price recognition results; Based on the identification results of the multiple advertising product types, cluster analysis is performed on the advertising database to obtain the first-level cluster identification results; Based on the price identification results of the multiple advertised products, cluster analysis is performed on the first-level cluster identification results to obtain the product advertising database.
3. The method of claim 1, wherein, The method for obtaining multiple product demand preference analysis results further includes: Obtain the preset set of historical time periods; Based on the preset historical time set, the consumption history information of the multiple users is collected to obtain the user consumption history database; Based on the user consumption history database, information on multiple consumer product types is obtained; Based on the information on multiple consumer product types, the user's consumption history database is statistically analyzed to obtain multiple consumption distribution statistical results. Based on the multiple consumption distribution statistics, the consumption popularity of the multiple consumer product types is evaluated to obtain multiple consumption popularity evaluation results; Based on the multiple consumer popularity assessment results, the information on the multiple consumer product types is labeled to obtain the analysis results of the multiple product demand preferences.
4. A network interconnection system for batch advertising, characterized by The system includes an advertising network interconnection module for executing the bulk advertising network interconnection method according to any one of claims 1 to 3, the system comprising: An advertising information collection module is used to collect basic information of multiple advertisements to obtain an advertising database. The identification and analysis module is used to perform feature identification on multiple advertisements based on the advertising database, obtain multiple advertising product feature identification results, and perform cluster analysis on the advertising database based on the multiple advertising product feature identification results to obtain a product advertising database. An advertising quality assessment module is used to assess the advertising quality of the multiple advertisements based on the product advertising database, and obtain multiple advertising quality assessment results. The product demand preference analysis module is used to build a user consumption history database, and to perform product demand preference analysis on multiple users based on the user consumption history database to obtain multiple product demand preference analysis results. An advertising user matching module is used to input the multiple advertising product feature identification results, the multiple advertising quality evaluation results, and the multiple product demand preference analysis results into an intelligent advertising user matching model to obtain multiple advertising user matching results. An advertising user interconnection module is used to interconnect the multiple advertisements and the multiple users through the advertising network interconnection module based on the multiple advertising user matching results and the product advertising database.