Advertisement putting content generation method and system based on large model

Abstract

The invention discloses an advertisement putting content generation method and system based on a large model, and relates to the technical field of advertisement putting content generation, and the advertisement putting content generation system comprises a data collection module, a data analysis module, a content generation module and a data verification module. Multi-channel sales related data is collected through a data collection module, user data and interaction information on social media are analyzed through a data analysis module, differential analysis should be carried out for demand information, behavioral habits and attitude tendencies of different channel users, and specific public opinion prevention or advertisement putting strategies are generated for different channels. And a set application effect is realized. Advanced data mining and artificial intelligence large model technologies are utilized, efficient brand advertisement contents adapted to different channels are automatically generated, the advertisement contents comprise long video delivery contents, short video delivery contents and image-text delivery contents, intelligence, individuation and dynamics of advertisement delivery are achieved, and the accuracy and effectiveness of brand promotion are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of advertisement delivery content generation, and in particular to a method and system for advertisement delivery content generation based on a large model. Background Art

[0002] The Chinese invention patent application with publication number CN118051631A discloses an information analysis management method and system for digital new media based on big data, including obtaining digital media data, obtaining digital media standard data based on digital media data based on data preprocessing, and obtaining digital media data feature information based on digital media classification database based on data mining. The present invention improves the efficiency and accuracy of subsequent keyword extraction by performing data preprocessing on digital media data, understands the demand information, behavioral habits and attitude tendencies of digital media users through data mining, monitors public opinion information related to the target on social media, news websites, forums and other media through public opinion monitoring keyword information, judges whether the public opinion attention is too high through the public opinion correlation index of digital media data, and adjusts the strategy in time through the public opinion comment index of digital media data, thereby improving the market feedback acceptance.

[0003] Problems with the above existing technologies: The analysis strategies for different channels are not clear: for different channels, the demand information, behavior habits and attitude tendencies of users should be analyzed differently. The application strategies for different channels are not clear: for different channels, specific public opinion prevention or advertising strategies can be generated to achieve the intended application effects. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method and system for generating advertisement content based on a large model. The following technical solutions are adopted:

[0005] A method for generating advertisement delivery content based on a large model, characterized by comprising the following steps:

[0006] Step 1: The data collection module collects marketing-related data from multiple channels. The marketing-related data includes social media brand-related data, consumer questionnaire results data, and historical data of previous marketing activities uploaded by the enterprise.

[0007] Step 2: The data analysis module analyzes user behavior and preferences on different channels to identify the content type and style that best suits each channel;

[0008] Step 3: The data analysis module uses data analysis technology to conduct in-depth analysis of user data in different regions, identify regional consumption characteristics and trends, and customize advertising strategies that conform to regional culture;

[0009] Step 4: The content generation module interacts with the data analysis module to develop advertising strategies and generates video marketing content and graphic marketing content based on the big model.

[0010] By adopting the above technical solutions, the data collection module collects multi-channel sales-related data, and the data analysis module analyzes user data and interactive information on social media. For different channels, the user's demand information, behavioral habits and attitude tendencies should be analyzed differently. For different channels, specific public opinion prevention or advertising strategies can be generated to achieve the established application effects. Using advanced data mining and artificial intelligence large model technology, efficient brand advertising content that adapts to different channels is automatically generated, realizing intelligent, personalized and dynamic advertising, and improving the accuracy and effectiveness of brand promotion.

[0011] Optionally, in step 1, the data collection module obtains social media data through API or RPA, and automatically extracts brand-related user comments, sharing, and likes data.

[0012] Optionally, the frequency of collecting social media data by the data collection module is set to once per hour.

[0013] By adopting the above technical solutions, the data collection module obtains social media data through API or RPA, and automatically extracts brand-related user comments, shares and likes data. The data collection frequency is set to once an hour to ensure the latest user interaction information is obtained.

[0014] Optionally, in step 1, the data collection module inputs the survey questions into the AI ​​big model, and the AI ​​big model automatically generates a user survey questionnaire, distributes the questionnaire through an online survey tool, collects consumers' preferences and expectations for the product, and directly integrates the questionnaire results into the data warehouse of the data collection module.

[0015] By adopting the above technical solutions, the AI ​​big model can be a big model that is fine-tuned based on the Qianwen open source big model. It can automatically generate user survey questionnaires based on survey questions and distribute questionnaires through online survey tools. Ultimately, it can collect consumers' preferences and expectations for products and provide a more direct user feedback channel.

[0016] Optionally, in step 1, the enterprise marketing team uploads historical data of previous marketing activities through the interface provided by the data collection module, including key performance indicator KPI data and customer feedback data. The file format supports CSV and Excel.

[0017] Optionally, the data analysis module uses the following two-stage analysis:

[0018] Data analysis by channel: Based on the questionnaire data organized in the data warehouse, dimensional modeling is performed to identify user interests and interaction patterns on different media platforms, and visual reports are generated to guide content creation.

[0019] Data analysis by region: Apply geographic information system analysis technology to divide user data by geographical location, analyze user preferences in different regions, and adjust advertising strategies based on the distribution of user groups.

[0020] Optionally, the content generation module deploys a large model, and the advertising strategy data output by the data analysis module is input into the large model. The large model generates marketing content in the following three ways:

[0021] Long video content generation: The big model generates advertising content in the long video format, including script writing and visual effects creativity; the big model analyzes existing long video content and finds advertising locations suitable for brand placement;

[0022] Short video content generation: The big model generates creative content suitable for short video platforms based on analysis results, focusing on high interactivity and rapid dissemination;

[0023] Graphic content generation: Use natural language processing and generation technology to automatically create advertising graphic content to support brand marketing activities on social media.

[0024] By adopting the above technical solutions, the big model can be self-developed or commercial AI big model can be used. For example, the intelligent marketing generation model (SMG) is combined with Adobe Premiere to produce high-quality long video delivery content. The video length is controlled within 3 to 5 minutes, which is suitable for delivery on platforms such as YouTube.

[0025] The intelligent marketing generation model (SMG) is used to generate long video advertising scripts, and short videos of 15 to 60 seconds are produced based on creative templates in the database, emphasizing visual impact and interactivity.

[0026] The intelligent marketing generation model (SMG) is used to generate graphic content, automatically generate creative advertising copy and illustrations, and is suitable for platforms such as WeChat public accounts.

[0027] Optionally, it also includes a data verification stage, in which the data verification module interacts with the data analysis module to generate advertising strategy results, designs questionnaires for small-scale sample confirmation, and collects targeted questionnaire questions by fully considering channel and regional differences to verify the accuracy of data analysis results.

[0028] A large-model-based advertising content generation system is used to implement a large-model-based advertising content generation method. The advertising content generation system includes a data acquisition module, a data analysis module, a content generation module and a data verification module. The data acquisition module is used to collect marketing-related data from multiple channels. The data analysis module is communicated with the data acquisition module and is used to analyze marketing-related data. The content generation module and the data verification module are respectively communicated with the data analysis module.

[0029] In summary, the present invention includes at least one of the following beneficial technical effects:

[0030] The present invention can provide a method and system for generating advertising content based on a large model. The data collection module collects multi-channel sales-related data, and the data analysis module analyzes user data and interactive information on social media. Differential analysis should be conducted for the demand information, behavioral habits, and attitude tendencies of users in different channels, and specific public opinion prevention or advertising strategies are generated for different channels to achieve the intended application effect. Advanced data mining and artificial intelligence large model technology are used to automatically generate efficient brand advertising content that adapts to different channels, realize intelligent, personalized, and dynamic advertising, and improve the accuracy and effectiveness of brand promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a flow chart of a method for generating advertisement delivery content based on a large model according to the present invention. DETAILED DESCRIPTION

[0032] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0033] The embodiment of the present invention discloses a method and system for generating advertisement delivery content based on a large model.

[0034] Reference Figure 1 A method for generating advertisement delivery content based on a large model, characterized by comprising the following steps:

[0035] Step 1: The data collection module collects marketing-related data from multiple channels. The marketing-related data includes social media brand-related data, consumer questionnaire results data, and historical data of previous marketing activities uploaded by the enterprise.

[0036] Step 2: The data analysis module analyzes user behavior and preferences on different channels to identify the content type and style that best suits each channel;

[0037] Step 3: The data analysis module uses data analysis technology to conduct in-depth analysis of user data in different regions, identify regional consumption characteristics and trends, and customize advertising strategies that conform to regional culture;

[0038] Step 4: The content generation module interacts with the data analysis module to develop advertising strategies and generates video marketing content and graphic marketing content based on the big model.

[0039] The data collection module collects multi-channel sales-related data, and the data analysis module analyzes user data and interactive information on social media. For different channels, the user's demand information, behavioral habits and attitude tendencies should be analyzed differently. For different channels, specific public opinion prevention or advertising strategies can be generated to achieve the established application effect. Using advanced data mining and artificial intelligence large model technology, efficient brand advertising content that adapts to different channels is automatically generated, realizing intelligent, personalized and dynamic advertising, and improving the accuracy and effectiveness of brand promotion.

[0040] In step 1, the data collection module obtains social media data through API or RPA, and automatically extracts brand-related user comments, shares, and likes data.

[0041] The frequency of collecting social media data in the data collection module is set to once per hour.

[0042] The data collection module obtains social media data through API or RPA, and automatically extracts brand-related user comments, shares, and likes data. The data collection frequency is set to once an hour to ensure the latest user interaction information is obtained.

[0043] In step 1, the data collection module inputs the survey questions into the AI ​​big model, and the AI ​​big model automatically generates a user survey questionnaire, distributes the questionnaire through online survey tools, collects consumers' preferences and expectations for products, and directly integrates the questionnaire results into the data warehouse of the data collection module.

[0044] The AI ​​big model can be a big model that has been fine-tuned based on the Qianwen open source big model. It can automatically generate user survey questionnaires based on survey questions and distribute the questionnaires through online survey tools. It can ultimately collect consumers' preferences and expectations for products and provide a more direct user feedback channel.

[0045] In step 1, the enterprise marketing team uploads historical data of previous marketing activities through the interface provided by the data collection module, including key performance indicator KPI data and customer feedback data. The file format supports CSV and Excel.

[0046] The data analysis module uses the following two-stage analysis:

[0047] Data analysis by channel: Based on the questionnaire data organized in the data warehouse, dimensional modeling is performed to identify user interests and interaction patterns on different media platforms, and visual reports are generated to guide content creation.

[0048] Data analysis by region: Apply geographic information system analysis technology to divide user data by geographical location, analyze user preferences in different regions, and adjust advertising strategies based on the distribution of user groups.

[0049] The content generation module deploys a big model, and the advertising strategy data output by the data analysis module is input into the big model. The big model generates marketing content in the following three ways:

[0050] Long video content generation: The big model generates advertising content in the long video format, including script writing and visual effects creativity; the big model analyzes existing long video content and finds advertising locations suitable for brand placement;

[0051] Short video content generation: The big model generates creative content suitable for short video platforms based on analysis results, focusing on high interactivity and rapid dissemination;

[0052] Graphic content generation: Use natural language processing and generation technology to automatically create advertising graphic content to support brand marketing activities on social media.

[0053] The big model can be self-developed or commercial AI big model can be used. For example, the Smart Marketing Generation Model (SMG) can be combined with Adobe Premiere to produce high-quality long video content. The video length is controlled within 3 to 5 minutes, which is suitable for delivery on platforms such as YouTube.

[0054] The intelligent marketing generation model (SMG) is used to generate long video advertising scripts, and short videos of 15 to 60 seconds are produced based on creative templates in the database, emphasizing visual impact and interactivity.

[0055] The intelligent marketing generation model (SMG) is used to generate graphic content, automatically generate creative advertising copy and illustrations, and is suitable for platforms such as WeChat public accounts.

[0056] It also includes the data verification stage, in which the data verification module and the data analysis module interact to generate advertising strategy results, design questionnaires for small-scale sample confirmation, and collect targeted questionnaire questions by fully considering channel and regional differences to verify the accuracy of data analysis results.

[0057] A large-model-based advertising content generation system is used to implement a large-model-based advertising content generation method. The advertising content generation system includes a data acquisition module, a data analysis module, a content generation module and a data verification module. The data acquisition module is used to collect marketing-related data from multiple channels. The data analysis module is communicated with the data acquisition module and is used to analyze marketing-related data. The content generation module and the data verification module are respectively communicated with the data analysis module.

[0058] The following is a specific case to illustrate the implementation principle of a method and system for generating advertising content based on a large model:

[0059] Client: A pharmaceutical company

[0060] Product: Oral liquid

[0061] Core issue: How to deliver targeted regional advertising content for a certain oral liquid.

[0062] Implementation process:

[0063] Data collection module: Social media data collection: Based on the product keywords (a certain oral liquid) and competitor keywords (Huoxiang Zhengqi Water, Shidishui), data is collected from the entire network, including Weibo, Toutiao, Xiaohongshu, Douyin, and Kuaishou. OCR technology is used in the collection process to obtain the keyword content that appears in the pictures and videos to ensure the comprehensiveness of the data.

[0064] Questionnaire generation and data collection: The specific method is: enter the product name and product mental characteristics, and use the big model to automatically generate 3-5 single-choice and multiple-choice questions. Note: The big model here is a fine-tuned big model based on the Qianwen open source big model, which has been injected with a lot of brand and medical knowledge. The generated questionnaire can be sent to the designated user through the mini program to fill in.

[0065] Enterprise data upload: supports enterprises to upload internal marketing materials. Enterprises usually have a large amount of approved marketing materials, so they can be used as material input for subsequent large model content generation.

[0066] Data Analysis Module:

[0067] Text analysis and topic extraction: NLP analysis is performed based on the acquired data, including word segmentation, stop word removal, intent identification, and topic LDA extraction. Through social media data and NLP analysis, we can clearly get the focus of this product and competitors in social media. Such as heatstroke prevention, travel essentials, travel partners, etc. We can also see the difference: a certain oral liquid will focus on alcohol-free, and it can be drunk while driving, while Huoxiang Zhengqi water contains alcohol.

[0068] Analysis by channel data: Based on data from different channels, such as Douyin and Xiaohongshu, differentiated comparative analysis is conducted. For example, Xiaohongshu focuses more on outdoor essentials, while Douyin focuses more on drinking while driving.

[0069] Analysis by time data: Spatial differences in thematic analysis based on product extraction. For example, different regions have different descriptions of heatstroke prevention based on a certain oral liquid, such as dry heatstroke in Central China and humid heatstroke in the Pearl River Delta.

[0070] Data Validation Module

[0071] Based on the above analysis, we further designed a questionnaire to conduct small-scale sample confirmation, and collected targeted questionnaire questions by fully considering channel and regional differences.

[0072] Content generation module

[0073] Long video: Collect all script data as input for the self-developed big model, and find suitable advertising positions based on the specific mindset of the brand. For example, if a certain oral liquid is alcohol-free and can be consumed while driving, a large number of scenes of driving vans and trucks can be found for advertising.

[0074] Short videos: Generate shooting scripts with regional characteristics based on the mindset and platform characteristics of users in short videos (TikTok, Kuaishou). For example, in Beijing, generate outdoor heatstroke prevention advertising scripts for scenic spots and tourism; in Zhejiang and Shanghai, generate heatstroke prevention advertising scripts for rental agencies and house hunting.

[0075] Graphic content: Generate graphic content with regional characteristics based on the mindset and platform characteristics of users (Weibo, Xiaohongshu). For example, in Guangdong, generate UGC content on damp heatstroke prevention; in Henan, generate UGC content on dry heatstroke prevention.

[0076] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for generating advertisement content based on a large model, characterized in that: The following steps are involved: Step 1: The data collection module collects marketing-related data from multiple channels. The marketing-related data includes social media brand-related data, consumer questionnaire results data, and historical data of previous marketing activities uploaded by the enterprise. Step 2: The data analysis module analyzes user behavior and preferences on different channels to identify the content type and style that best suits each channel; Step 3: The data analysis module uses data analysis technology to conduct in-depth analysis of user data in different regions, identify regional consumption characteristics and trends, and customize advertising strategies that conform to regional culture; Step 4: The content generation module interacts with the data analysis module to develop advertising strategies and generates video marketing content and graphic marketing content based on the big model.

2. The method for generating advertisement content based on a large model according to claim 1, characterized in that: In step 1, the data collection module obtains social media data through API or RPA, and automatically extracts brand-related user comments, shares, and likes data.

3. The method for generating advertisement content based on a large model according to claim 2, characterized in that: The frequency of collecting social media data in the data collection module is set to once per hour.

4. The method for generating advertisement content based on a large model according to claim 3, characterized in that: In step 1, the data collection module inputs the survey questions into the AI ​​big model, and the AI ​​big model automatically generates a user survey questionnaire, distributes the questionnaire through online survey tools, collects consumers' preferences and expectations for products, and directly integrates the questionnaire results into the data warehouse of the data collection module.

5. The method for generating advertisement content based on a large model according to claim 4, characterized in that: In step 1, the enterprise marketing team uploads historical data of previous marketing activities through the interface provided by the data collection module, including key performance indicator KPI data and customer feedback data. The file format supports CSV and Excel.

6. The method for generating advertisement delivery content based on a large model according to claim 5, characterized in that: The data analysis module uses the following two-stage analysis: Data analysis by channel: Based on the questionnaire data organized in the data warehouse, dimensional modeling is performed to identify user interests and interaction patterns on different media platforms, and visual reports are generated to guide content creation; Data analysis by region: Apply geographic information system analysis technology to divide user data by geographical location, analyze user preferences in different regions, and adjust advertising strategies based on the distribution of user groups.

7. The method for generating advertisement content based on a large model according to claim 6, characterized in that: The content generation module deploys a big model, and the advertising strategy data output by the data analysis module is input into the big model. The big model generates marketing content in the following three ways: Long video content generation: The big model generates advertising content in the long video format, including script writing and visual effects creativity; the big model analyzes existing long video content and finds advertising locations suitable for brand placement; Short video content generation: The big model generates creative content suitable for short video platforms based on analysis results, focusing on high interactivity and rapid dissemination; Graphic content generation: Use natural language processing and generation technology to automatically create advertising graphic content to support brand marketing activities on social media.

8. The method for generating advertisement content based on a large model according to claim 7, characterized in that: It also includes the data verification stage, in which the data verification module and the data analysis module interact to generate advertising strategy results, design questionnaires for small-scale sample confirmation, and collect targeted questionnaire questions by fully considering channel and regional differences to verify the accuracy of data analysis results.

9. A system for generating advertisement content based on a large model, characterized in that: A method for generating advertising content based on a large model according to claim 8 is used to implement the advertising content generation system, which includes a data acquisition module, a data analysis module, a content generation module and a data verification module. The data acquisition module is used to collect marketing-related data from multiple channels. The data analysis module is communicated with the data acquisition module and is used to analyze marketing-related data. The content generation module and the data verification module are respectively communicated with the data analysis module.

Images