Artificial intelligence marketing method and system based on big data model

Abstract

The invention relates to the related field of marketing management, and discloses an artificial intelligence marketing method and system based on a big data model, intelligent object-based deep customization marketing promotion is realized through big data, matching is performed according to a big data portrait of a user and a theme required by promotion, and the marketing promotion efficiency is improved. According to the method, the subject content with high interest deviation of the promotion object is obtained, and the marketing promotion content matched with the promotion object one by one is correspondingly generated, so that the promotion effect is optimized, and the actual proportion of effective promotion is improved.

Description

Technical Field

[0001] The present invention relates to fields related to marketing management, and specifically to an artificial intelligence marketing method and system based on a big data model. Background Art

[0002] In the online marketing implementation of the existing technology, the marketing advertisements are mostly placed at fixed points on the website, that is, marketing advertisements with fixed content are pushed through fixed advertising positions. Such a wide-ranging marketing method is used to tap into potential users and achieve the purpose of marketing promotion.

[0003] However, with this type of marketing promotion method, different users see the same marketing content at the same point in time, and ultimately, the content obtained by the user group that receives the same marketing promotion is the same. This type of marketing promotion method is limited by the differences in user groups, and the actual marketing effect is extremely low. Summary of the invention

[0004] The purpose of the present invention is to provide an artificial intelligence marketing method and system based on a big data model to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An artificial intelligence marketing method based on a big data model, comprising:

[0007] Generate a big data portrait of the promotion object; obtain the historical browsing records of the promotion object on several sites, collect statistics on the browsing records of the promotion object based on the browsing content characteristics, and generate a big data portrait of the promotion object;

[0008] Acquisition of promotion subjects and determination of themes; Matching and selecting the content to be promoted for execution from a number of marketing objects, and acquiring promotion replacement themes for the content to be promoted, wherein the promotion replacement themes are modular media objects containing an indefinite set of replaceable data, and the replaceable data are provided with replacement content features;

[0009] Based on the theme search and matching of the promotion object; search the big data portrait through the content replacement feature, and select the highly matched replaceable data based on the matching degree sequence obtained by the search;

[0010] Customized update and push of the promotion subject's theme; update the promotion replacement theme through replaceable data to generate deeply customized content to be promoted that matches the promotion object, and output the content to be promoted through the promotion object end.

[0011] As a further solution of the present invention: the big data portrait is used to characterize the browsing preference statistical model of the promotion object based on the browsing content characteristics, the browsing preference statistical model is a tree-like distribution structure with multiple feature levels, and the multiple feature levels are respectively used to characterize the multiple guiding classification feature tags of the user's browsing content, and the multiple classification feature tags are progressively associated or parallel associated. The acquisition of the historical browsing records of the promotion object on several sites mainly includes:

[0012] Acquisition and evaluation of site browsing data records, the site browsing data including browsing records, click records, shopping records, browsing completion rates corresponding to browsing objects, and user characteristic duration ratios relative to average browsing durations of browsing objects, the browsing completion rates and user characteristic duration ratios are used to characterize the user's interest preference values ​​for the corresponding browsing objects;

[0013] The consumption data records of the promotion objects are acquired and evaluated, wherein the consumption records include the types of consumer goods and the corresponding value ranges of the consumer goods among the same type of products. The types of consumer goods and the value ranges are used to represent the consumption preferences and consumption capabilities of the users, respectively.

[0014] As a further solution of the present invention: the step of matching and selecting the promotion contents to be executed for a plurality of marketing objects specifically includes:

[0015] Based on the classification labels related to the interest preferences of the promotion objects in the big data portrait, a plurality of the marketing objects are matched and screened to obtain a plurality of marketing objects that meet the preferences;

[0016] Based on the consumption preferences and consumption capabilities of the promotion targets in the big data portrait, several marketing targets that meet the preferences are further screened to obtain content to be promoted, which meets the interest preferences and consumption capability preferences of the promotion targets.

[0017] As a further solution of the present invention: the content to be promoted includes a promotion subject and a promotion replacement subject;

[0018] The promotion subject is used to represent the promotion information framework of the content to be promoted, and the promotion information framework may include an image information framework and a streaming media information framework, which are respectively used to represent the content promotion implementation method of the content to be promoted;

[0019] The promotion replacement theme is used to characterize the customized replacement content segments in the image information framework and the streaming media information framework. The promotion replacement theme is provided with an information distribution specification. The information distribution specification is used to characterize the embedding execution mode and characterization mode of the replaceable data obtained through the promotion object. The embedding execution mode is used to characterize the scaling and embedding angle of the replaceable data. The characterization mode is used to characterize the characterization scheme of the display effect of the replaceable data, so as to achieve style consistency with the promotion.

[0020] As a further solution of the present invention, it also includes a permission verification step for obtaining replaceable data:

[0021] Performing content recognition on the replaceable data with a high matching degree obtained based on the search, obtaining corresponding authority information, and validating the replaceable data if the authority information is in an available state, wherein the validation is used to represent the available state of the promotion replacement theme;

[0022] If the permission information is in an unavailable state, the permission information is determined for multiple replaceable data that reach a preset matching rate based on the matching degree sequence order, and the replaceable data is selected and validated accordingly;

[0023] If multiple replaceable data that reach a preset matching rate are all characterized as a state where permission information is unavailable, feature labels are extracted through the replaceable data with the highest matching rate, and AI content creation is performed based on the feature labels to generate replaceable data without permission information.

[0024] The embodiment of the present invention aims to provide an artificial intelligence marketing system based on a big data model, comprising:

[0025] The object portrait acquisition module is used to generate a big data portrait of the promotion object; obtain the historical browsing records of the promotion object on several sites, collect statistics on the browsing records of the promotion object based on the browsing content characteristics, and generate a big data portrait of the promotion object;

[0026] The promotion demand acquisition module is used to acquire the promotion subject and determine the theme; match and select the content to be promoted for execution from several marketing objects, and acquire the promotion replacement theme of the content to be promoted, wherein the promotion replacement theme is a modular media object containing an indefinite group of replaceable data, and the replaceable data is provided with replacement content features;

[0027] Theme data matching module is used for theme retrieval and matching based on the promotion object; it searches the big data portrait through the content replacement feature, and selects the highly matched replaceable data based on the matching degree sequence obtained by the search;

[0028] The customized update push module is used for the customized update and push of the theme of the promotion subject; the promotion replacement theme is updated through replaceable data to generate deeply customized content to be promoted that matches the promotion object, and the content to be promoted is output through the promotion object end.

[0029] As a further solution of the present invention: the big data portrait is used to characterize the browsing preference statistical model of the promotion object based on the browsing content characteristics, the browsing preference statistical model is a tree-like distribution structure with multiple feature levels, and the multiple feature levels are respectively used to characterize the multiple guiding classification feature tags of the user's browsing content, and the multiple classification feature tags are progressively associated or parallel associated. The acquisition of the historical browsing records of the promotion object on several sites mainly includes:

[0030] Acquisition and evaluation of site browsing data records, the site browsing data including browsing records, click records, shopping records, browsing completion rates corresponding to browsing objects, and user characteristic duration ratios relative to average browsing durations of browsing objects, the browsing completion rates and user characteristic duration ratios are used to characterize the user's interest preference values ​​for the corresponding browsing objects;

[0031] The consumption data records of the promotion objects are acquired and evaluated, wherein the consumption records include the types of consumer goods and the corresponding value ranges of the consumer goods among the same type of products. The types of consumer goods and the value ranges are used to represent the consumption preferences and consumption capabilities of the users, respectively.

[0032] As a further solution of the present invention: the promotion demand acquisition module includes:

[0033] A preference matching unit is used to match and screen a plurality of marketing objects based on the classification tags related to the interest preferences of the promotion objects in the big data portrait, and obtain a plurality of marketing objects that meet the preferences;

[0034] The capability matching unit is used to further screen several marketing objects that meet the preferences based on the consumption preferences and consumption capabilities of the promotion objects in the big data portrait, so as to obtain content to be promoted, and the content to be promoted meets the interest preferences and consumption capability preferences of the promotion objects.

[0035] As a further solution of the present invention: the content to be promoted includes a promotion subject and a promotion replacement subject;

[0036] The promotion subject is used to represent the promotion information framework of the content to be promoted, and the promotion information framework may include an image information framework and a streaming media information framework, which are respectively used to represent the content promotion implementation method of the content to be promoted;

[0037] The promotion replacement theme is used to characterize the customized replacement content segments in the image information framework and the streaming media information framework. The promotion replacement theme is provided with an information distribution specification. The information distribution specification is used to characterize the embedding execution mode and characterization mode of the replaceable data obtained through the promotion object. The embedding execution mode is used to characterize the scaling and embedding angle of the replaceable data. The characterization mode is used to characterize the characterization scheme of the display effect of the replaceable data, so as to achieve style consistency with the promotion.

[0038] As a further solution of the present invention: it also includes replacing the authority verification module, specifically including:

[0039] The authority determination unit is used to perform content recognition on the replaceable data with a high matching degree obtained based on the search, obtain corresponding authority information, and if the authority information is in an available state, validate the replaceable data, wherein the validation is used to represent the available state of the promotion replacement theme;

[0040] A sequence determination unit, configured to determine the permission information of a plurality of replaceable data that reach a preset matching rate based on a matching degree sequence order if the permission information is in an unavailable state, and select and validate the replaceable data accordingly;

[0041] The AI ​​auxiliary unit is used to extract feature labels through the replaceable data with the highest matching rate if multiple replaceable data that reach a preset matching rate are all characterized as a state where permission information is unavailable, and to create AI content based on the feature labels to generate replaceable data without permission information.

[0042] Compared with the prior art, the beneficial effects of the present invention are: realizing intelligent object-based deeply customized marketing promotion through big data, matching the user's big data portrait with the theme required for promotion, obtaining the theme content with higher interest preference of the promotion object, and generating corresponding marketing promotion content that matches the promotion object one by one, thereby optimizing the promotion effect and increasing the actual proportion of effective promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 The figure is a flowchart of an artificial intelligence marketing method based on a big data model.

[0044] Figure 2 The present invention is a flowchart of the steps of preparing to execute content to be promoted in an artificial intelligence marketing method based on a big data model.

[0045] Figure 3 This is a block diagram of the composition of an artificial intelligence marketing system based on a big data model. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0047] The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.

[0048] like Figure 1 An artificial intelligence marketing method based on a big data model provided by an embodiment of the present invention comprises the following steps:

[0049] S10, generating a big data portrait of the promotion object; obtaining the historical browsing records of the promotion object on several sites, collecting statistics on the browsing records of the promotion object based on the browsing content characteristics, and generating a big data portrait of the promotion object;

[0050] S20, obtaining the promotion subject and determining the theme; matching and selecting the to-be-promoted content to be executed for a plurality of marketing objects, and obtaining the promotion replacement theme of the to-be-promoted content, wherein the promotion replacement theme is a modular media object containing an indefinite group of replaceable data, and the replaceable data is provided with replacement content features;

[0051] S30, based on the theme search and matching of the promotion object; searching the big data portrait through the content replacement feature, and selecting the replaceable data with high matching degree based on the matching degree sequence obtained by the search;

[0052] S40, customized update and push of the theme of the promotion subject; the promotion replacement theme is updated through replaceable data to generate deeply customized content to be promoted that matches the promotion object, and the content to be promoted is output through the promotion object end.

[0053] In this embodiment, an artificial intelligence marketing method based on a big data model is provided, which is used to realize intelligent object-based deep customized marketing promotion through big data, match the user's big data portrait with the theme required for promotion, obtain the theme content with higher interest preference of the promotion object, and generate marketing promotion content that matches the promotion object one by one, thereby optimizing the promotion effect and increasing the actual proportion of effective promotion; in the prior art, online promotion of marketing is mostly carried out in a fixed-point delivery manner, that is, when the user browses the content, the content is delivered to a certain advertising node, so the marketing content viewed by different users at the same node may be the same (or there are multiple in the library, randomly or according to user preferences, but the content remains unchanged and the same), this method has the problem of low actual marketing effect, because different people have different content preferences for product types, etc., so the traditional marketing delivery method is an inefficient wide-net solution; the specific implementation process of this application can be summarized as follows: the promotion object corresponds to a large number of advertisements obtained through historical browsing record statistics According to the portrait, the preferences of the promotion object in various fields are expressed. For example, if the user prefers to browse military media, landscape media, etc., then the user will have corresponding high-frequency browsing objects, such as a certain type of military equipment or a specific model of fighter jet, etc.; and at the advertising marketing node of a certain website, there are multiple advertisements waiting for promotion, all of which are the contractual objects with the owner of the node, and each content of the promotion, whether it is an image marketing advertisement or a video media marketing advertisement, includes an irreplaceable promotion subject and a customizable promotion replacement theme. Taking image marketing advertisements as an example, the entire picture including the main text introduction information and other content are the promotion themes, and some of the images and other content involved can be set as promotion replacement themes, that is, the theme content of the advertisement, such as mobile phone cases and other related products, whose pattern content can be customized based on the preferences of the promotion object. Taking the aforementioned content as an example, a customized promotion replacement theme can be generated according to the military equipment image preferred by the object to replace the printed image of the mobile phone case in the promotion content, so as to push customized marketing content.

[0054] As another preferred embodiment of the present invention, the big data portrait is used to characterize the browsing preference statistical model of the promotion object based on the browsing content characteristics, the browsing preference statistical model is a tree-like distribution structure with multiple feature levels, and the multiple feature levels are respectively used to characterize the multiple guiding classification feature tags of the user's browsing content, and the multiple classification feature tags are progressively associated or parallel associated. The acquisition of the historical browsing records of the promotion object on several sites mainly includes:

[0055] Acquisition and evaluation of site browsing data records, the site browsing data including browsing records, click records, shopping records, browsing completion rates corresponding to browsing objects, and user characteristic duration ratios relative to average browsing durations of browsing objects, the browsing completion rates and user characteristic duration ratios are used to characterize the user's interest preference values ​​for the corresponding browsing objects;

[0056] The consumption data records of the promotion objects are acquired and evaluated, wherein the consumption records include the types of consumer goods and the corresponding value ranges of the consumer goods among the same type of products. The types of consumer goods and the value ranges are used to represent the consumption preferences and consumption capabilities of the users, respectively.

[0057] In this embodiment, the big data portrait is further explained, wherein the browsing completion rate and the user characteristic time ratio are both used to characterize the user's preference degree. For example, for the browsing record of a certain media video, the higher the user's browsing completion rate, the higher the user's preference degree is. The larger the ratio of the browsing time length to the actual time required (or the average reading time, which depends on the media type. For video types, the actual time required to complete the browsing is certain; for text and image types, the average time of other users is taken), the higher the interest in the object is. The value range is used to indicate the user's consumption ability, or more accurately expressed as consumption preference, and for different types of items, the range of money that the user will spend on consumption. This range corresponds to the value span of the same type of goods.

[0058] like Figure 2 As shown, as another preferred embodiment of the present invention, the step of matching and selecting the promotion contents to be executed for several marketing objects specifically includes:

[0059] S21, matching and screening a plurality of marketing objects based on the classification labels related to the interest preferences of the promotion objects in the big data portrait, and obtaining a plurality of marketing objects that meet the preferences;

[0060] S22, further screening several marketing objects that meet the preferences based on the consumption preferences and consumption capabilities of the promotion objects in the big data portrait to obtain content to be promoted, wherein the content to be promoted meets the interest preferences and consumption capability preferences of the promotion objects.

[0061] In this embodiment, the steps of obtaining the promotion subject and judging the theme are supplemented, mainly to screen several contents to be promoted of the current advertising promotion node according to the big data portrait conditions of the promotion object, and select marketing content suitable for the current user object. The first one includes type preference. For example, if the current object is a male object, women's products are filtered to match the product field with high frequency preferences of the object; the second one is the screening of consumption preferences, that is, the price range of marketing products is screened according to the user's consumption range preference, so as to match marketing content that is more suitable for the object and improve the marketing success rate.

[0062] As another preferred embodiment of the present invention, the content to be promoted includes a promotion subject and a promotion replacement subject;

[0063] The promotion subject is used to represent the promotion information framework of the content to be promoted, and the promotion information framework may include an image information framework and a streaming media information framework, which are respectively used to represent the content promotion implementation method of the content to be promoted;

[0064] The promotion replacement theme is used to characterize the customized replacement content segments in the image information framework and the streaming media information framework. The promotion replacement theme is provided with an information distribution specification. The information distribution specification is used to characterize the embedding execution mode and characterization mode of the replaceable data obtained through the promotion object. The embedding execution mode is used to characterize the scaling and embedding angle of the replaceable data. The characterization mode is used to characterize the characterization scheme of the display effect of the replaceable data, so as to achieve style consistency with the promotion.

[0065] In this embodiment, the promotion subject and the promotion replacement theme are explained in detail. It can be simply understood that the promotion subject is the main framework content of the advertisement, such as a video or an advertising picture, and the promotion replacement theme is used to represent the characteristic content in this video or advertisement, such as image content, and optional features of the product (different colors or appearance options), etc.

[0066] As another preferred embodiment of the present invention, it also includes a permission verification step for obtaining replaceable data:

[0067] Performing content recognition on the replaceable data with a high matching degree obtained based on the search, obtaining corresponding authority information, and validating the replaceable data if the authority information is in an available state, wherein the validation is used to represent the available state of the promotion replacement theme;

[0068] If the permission information is in an unavailable state, the permission information is determined for multiple replaceable data that reach a preset matching rate based on the matching degree sequence order, and the replaceable data is selected and validated accordingly;

[0069] If multiple replaceable data that reach a preset matching rate are all characterized as a state where permission information is unavailable, feature labels are extracted through the replaceable data with the highest matching rate, and AI content creation is performed based on the feature labels to generate replaceable data without permission information.

[0070] In this embodiment, a verification process for replaceable data is supplemented. Since this application is implemented based on the feature replacement of promotional content established based on the user's web browsing preferences, there will be a problem of the permissions of the feature data content obtained based on the user's big data portrait, that is, whether the corresponding content is open permission content that can be used directly. If not, new content is generated based on AI by obtaining its feature tags to avoid the misuse of unauthorized content and ensure the formalization of promotion.

[0071] like Figure 3 As shown, the present invention also provides an artificial intelligence marketing system based on a big data model, which includes:

[0072] The object portrait acquisition module 100 is used to generate a big data portrait of the promotion object; obtain the historical browsing records of the promotion object on several sites, collect statistics on the browsing records of the promotion object based on the browsing content characteristics, and generate a big data portrait of the promotion object;

[0073] The promotion demand acquisition module 200 is used to acquire the promotion subject and determine the theme; select the promotion content to be executed for matching with several marketing objects, and acquire the promotion replacement theme of the promotion content, wherein the promotion replacement theme is a modular media object containing an indefinite group of replaceable data, and the replaceable data is provided with replacement content features;

[0074] The subject data matching module 300 is used for subject search and matching based on the promotion object; searching the big data portrait through the content replacement feature, and selecting high matching replaceable data based on the matching degree sequence obtained by the search;

[0075] The customized update push module 400 is used for customized update and push of the theme of the promotion subject; the promotion replacement theme is updated through replaceable data to generate deeply customized content to be promoted that matches the promotion object, and the content to be promoted is output through the promotion object end.

[0076] As another preferred embodiment of the present invention, the big data portrait is used to characterize the browsing preference statistical model of the promotion object based on the browsing content characteristics, the browsing preference statistical model is a tree-like distribution structure with multiple feature levels, and the multiple feature levels are respectively used to characterize the multiple guiding classification feature tags of the user's browsing content, and the multiple classification feature tags are progressively associated or parallel associated. The acquisition of the historical browsing records of the promotion object on several sites mainly includes:

[0077] Acquisition and evaluation of site browsing data records, the site browsing data including browsing records, click records, shopping records, browsing completion rates corresponding to browsing objects, and user characteristic duration ratios relative to average browsing durations of browsing objects, the browsing completion rates and user characteristic duration ratios are used to characterize the user's interest preference values ​​for the corresponding browsing objects;

[0078] The consumption data records of the promotion objects are acquired and evaluated, wherein the consumption records include the types of consumer goods and the corresponding value ranges of the consumer goods among the same type of products. The types of consumer goods and the value ranges are used to represent the consumption preferences and consumption capabilities of the users, respectively.

[0079] As another preferred embodiment of the present invention, the promotion demand acquisition module includes:

[0080] A preference matching unit is used to match and screen a plurality of marketing objects based on the classification tags related to the interest preferences of the promotion objects in the big data portrait, and obtain a plurality of marketing objects that meet the preferences;

[0081] The capability matching unit is used to further screen several marketing objects that meet the preferences based on the consumption preferences and consumption capabilities of the promotion objects in the big data portrait, so as to obtain content to be promoted, and the content to be promoted meets the interest preferences and consumption capability preferences of the promotion objects.

[0082] As another preferred embodiment of the present invention, the content to be promoted includes a promotion subject and a promotion replacement subject;

[0083] The promotion subject is used to represent the promotion information framework of the content to be promoted, and the promotion information framework may include an image information framework and a streaming media information framework, which are respectively used to represent the content promotion implementation method of the content to be promoted;

[0084] The promotion replacement theme is used to characterize the customized replacement content segments in the image information framework and the streaming media information framework. The promotion replacement theme is provided with an information distribution specification. The information distribution specification is used to characterize the embedding execution mode and characterization mode of the replaceable data obtained through the promotion object. The embedding execution mode is used to characterize the scaling and embedding angle of the replaceable data. The characterization mode is used to characterize the characterization scheme of the display effect of the replaceable data, so as to achieve style consistency with the promotion.

[0085] As another preferred embodiment of the present invention, it also includes a replacement authority verification module, specifically including:

[0086] The authority determination unit is used to perform content recognition on the replaceable data with a high matching degree obtained based on the search, obtain corresponding authority information, and if the authority information is in an available state, validate the replaceable data, wherein the validation is used to represent the available state of the promotion replacement theme;

[0087] A sequence determination unit, configured to determine the permission information of a plurality of replaceable data that reach a preset matching rate based on a matching degree sequence order if the permission information is in an unavailable state, and select and validate the replaceable data accordingly;

[0088] The AI ​​auxiliary unit is used to extract feature labels through the replaceable data with the highest matching rate if multiple replaceable data that reach a preset matching rate are all characterized as a state where permission information is unavailable, and to create AI content based on the feature labels to generate replaceable data without permission information.

[0089] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0090] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the disclosure in the specification and examples. This application is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.

[0091] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. An artificial intelligence marketing method based on a big data model, characterized in that: Include: Generate a big data portrait of the promotion object; obtain the historical browsing records of the promotion object on several sites, collect statistics on the browsing records of the promotion object based on the browsing content characteristics, and generate a big data portrait of the promotion object; Acquisition of promotion subjects and determination of themes; Matching and selecting the content to be promoted for execution from a number of marketing objects, and acquiring promotion replacement themes for the content to be promoted, wherein the promotion replacement themes are modular media objects containing an indefinite set of replaceable data, and the replaceable data are provided with replacement content features; Based on the theme search and matching of the promotion object; search the big data portrait through the content replacement feature, and select the highly matched replaceable data based on the matching degree sequence obtained by the search; Customized update and push of the promotion subject's theme; updating the promotion replacement theme through replaceable data to generate deeply customized content to be promoted that matches the promotion object, and outputting the content to be promoted through the promotion object end.

2. The artificial intelligence marketing method based on big data model according to claim 1 is characterized in that: The big data portrait is used to characterize the browsing preference statistical model of the promotion object based on the browsing content characteristics. The browsing preference statistical model is a tree-like distribution structure with multiple feature levels. The multiple feature levels are respectively used to characterize multiple guiding classification feature tags of the user's browsing content. The multiple classification feature tags are progressively associated or parallel associated. The acquisition of the historical browsing records of the promotion object on several sites mainly includes: Acquisition and evaluation of site browsing data records, the site browsing data including browsing records, click records, shopping records, browsing completion rates corresponding to browsing objects, and user characteristic duration ratios relative to average browsing durations of browsing objects, the browsing completion rates and user characteristic duration ratios are used to characterize the user's interest preference values ​​for the corresponding browsing objects; The consumption data records of the promotion objects are acquired and evaluated, wherein the consumption records include the types of consumer goods and the corresponding value ranges of the consumer goods among the same type of products. The types of consumer goods and the value ranges are used to represent the consumption preferences and consumption capabilities of the users, respectively.

3. The artificial intelligence marketing method based on big data model according to claim 2 is characterized in that: The step of matching and selecting the promotion content to be executed for a plurality of marketing objects specifically includes: Based on the classification labels related to the interest preferences of the promotion objects in the big data portrait, a plurality of the marketing objects are matched and screened to obtain a plurality of marketing objects that meet the preferences; Based on the consumption preferences and consumption capabilities of the promotion targets in the big data portrait, several marketing targets that meet the preferences are further screened to obtain content to be promoted, which meets the interest preferences and consumption capability preferences of the promotion targets.

4. The artificial intelligence marketing method based on big data model according to claim 1 is characterized in that: The content to be promoted includes the promotion subject and the promotion replacement subject; The promotion subject is used to represent the promotion information framework of the content to be promoted, and the promotion information framework may include an image information framework and a streaming media information framework, which are respectively used to represent the content promotion implementation method of the content to be promoted; The promotion replacement theme is used to characterize the customized replacement content segments in the image information framework and the streaming media information framework. The promotion replacement theme is provided with an information distribution specification. The information distribution specification is used to characterize the embedding execution mode and characterization mode of the replaceable data obtained through the promotion object. The embedding execution mode is used to characterize the scaling and embedding angle of the replaceable data. The characterization mode is used to characterize the characterization scheme of the display effect of the replaceable data, so as to achieve style consistency with the promotion.

5. The artificial intelligence marketing method based on big data model according to claim 1 is characterized in that: It also includes the permission verification step to obtain replaceable data: Performing content recognition on the replaceable data with a high matching degree obtained based on the search, obtaining corresponding authority information, and validating the replaceable data if the authority information is in an available state, wherein the validation is used to represent the available state of the promotion replacement theme; If the permission information is in an unavailable state, the permission information is determined for multiple replaceable data that reach a preset matching rate based on the matching degree sequence order, and the replaceable data is selected and validated accordingly; If multiple replaceable data that reach a preset matching rate are all characterized as a state where permission information is unavailable, feature labels are extracted through the replaceable data with the highest matching rate, and AI content creation is performed based on the feature labels to generate replaceable data without permission information.

6. An artificial intelligence marketing system based on a big data model, characterized in that: Include: The object portrait acquisition module is used to generate a big data portrait of the promotion object; obtain the historical browsing records of the promotion object on several sites, collect statistics on the browsing records of the promotion object based on the browsing content characteristics, and generate a big data portrait of the promotion object; The promotion demand acquisition module is used to acquire the promotion subject and determine the theme; match and select the content to be promoted for execution from several marketing objects, and acquire the promotion replacement theme of the content to be promoted, wherein the promotion replacement theme is a modular media object containing an indefinite group of replaceable data, and the replaceable data is provided with replacement content features; Theme data matching module is used for theme retrieval and matching based on the promotion object; it searches the big data portrait through the content replacement feature, and selects the highly matched replaceable data based on the matching degree sequence obtained by the search; The customized update push module is used for the customized update and push of the theme of the promotion subject; the promotion replacement theme is updated through replaceable data to generate deeply customized content to be promoted that matches the promotion object, and the content to be promoted is output through the promotion object end.

7. The artificial intelligence marketing system based on big data model according to claim 6 is characterized in that: The big data portrait is used to characterize the browsing preference statistical model of the promotion object based on the browsing content characteristics. The browsing preference statistical model is a tree-like distribution structure with multiple feature levels. The multiple feature levels are respectively used to characterize multiple guiding classification feature tags of the user's browsing content. The multiple classification feature tags are progressively associated or parallel associated. The acquisition of the historical browsing records of the promotion object on several sites mainly includes: Acquisition and evaluation of site browsing data records, the site browsing data including browsing records, click records, shopping records, browsing completion rates corresponding to browsing objects, and user characteristic duration ratios relative to average browsing durations of browsing objects, the browsing completion rates and user characteristic duration ratios are used to characterize the user's interest preference values ​​for the corresponding browsing objects; The consumption data records of the promotion objects are acquired and evaluated, wherein the consumption records include the types of consumer goods and the corresponding value ranges of the consumer goods among the same type of products. The types of consumer goods and the value ranges are used to represent the consumption preferences and consumption capabilities of the users, respectively.

8. The artificial intelligence marketing system based on big data model according to claim 7 is characterized in that: The promotion demand acquisition module includes: A preference matching unit is used to match and screen a plurality of marketing objects based on the classification tags related to the interest preferences of the promotion objects in the big data portrait, and obtain a plurality of marketing objects that meet the preferences; The capability matching unit is used to further screen several marketing objects that meet the preferences based on the consumption preferences and consumption capabilities of the promotion objects in the big data portrait, so as to obtain content to be promoted, and the content to be promoted meets the interest preferences and consumption capability preferences of the promotion objects.

9. The artificial intelligence marketing system based on big data model according to claim 6 is characterized in that: The content to be promoted includes the promotion subject and the promotion replacement subject; The promotion subject is used to represent the promotion information framework of the content to be promoted, and the promotion information framework may include an image information framework and a streaming media information framework, which are respectively used to represent the content promotion implementation method of the content to be promoted; The promotion replacement theme is used to characterize the customized replacement content segments in the image information framework and the streaming media information framework. The promotion replacement theme is provided with an information distribution specification. The information distribution specification is used to characterize the embedding execution mode and characterization mode of the replaceable data obtained through the promotion object. The embedding execution mode is used to characterize the scaling and embedding angle of the replaceable data. The characterization mode is used to characterize the characterization scheme of the display effect of the replaceable data, so as to achieve style consistency with the promotion.

10. The artificial intelligence marketing system based on big data model according to claim 6, characterized in that: It also includes replacing the permission verification module, including: The authority determination unit is used to perform content recognition on the replaceable data with a high matching degree obtained based on the search, obtain corresponding authority information, and if the authority information is in an available state, validate the replaceable data, wherein the validation is used to represent the available state of the promotion replacement theme; A sequence determination unit, configured to determine the permission information of a plurality of replaceable data that reach a preset matching rate based on a matching degree sequence order if the permission information is in an unavailable state, and select and validate the replaceable data accordingly; The AI ​​auxiliary unit is used to extract feature labels through the replaceable data with the highest matching rate if multiple replaceable data that reach a preset matching rate are all characterized as a state where permission information is unavailable, and to create AI content based on the feature labels to generate replaceable data without permission information.

Images