Advertisement content customization generation method and system based on user portrait

Abstract

The application provides an advertisement content customization generation method and system based on a user portrait. The demand core elements are disassembled according to the behavior preference track and the demand potential tendency in the user portrait, and a demand gene sequence containing multiple evolvable demand fragments and associated rules is generated. Then, based on the evolvability of the demand fragments, an AI algorithm is called to generate multiple groups of advertisement content basic variants carrying gene identifiers. The basic variants are subjected to evolutionary iteration processing to generate multiple generations of evolutionary variants and evolutionary sequences by adjusting the text structure. The user interaction feedback information is collected to identify the variant with the highest adaptation degree, and the optimal evolution path is determined. The variant is optimized along the optimal path, and the core adaptation features are integrated to generate a customized advertisement content set that deeply matches the demand genes of the user portrait, thereby improving the advertisement accuracy and conversion rate.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and more specifically, to a method and system for customizing and generating advertising content based on user profiles. Background Technology

[0002] In today's digital marketing era, the accuracy and effectiveness of advertising are crucial for a company's marketing and business growth. Traditional advertising content generation methods are mainly based on market research and experience-based judgment. The advertising production team designs universally applicable advertising content based on a rough understanding of the target market, and then distributes it widely through various channels. However, the above methods have many drawbacks.

[0003] On the one hand, due to a lack of in-depth and detailed understanding of individual users, advertising content often fails to accurately match users' actual needs and interests, resulting in low ad appeal and conversion rates. For example, different users may have significantly different focuses on the same type of product; some users prioritize product performance, while others focus more on price and appearance. Traditional advertising struggles to meet these diverse needs. On the other hand, traditional advertising content lacks effective feedback mechanisms and dynamic adjustment capabilities after generation. Once an ad is launched, it's difficult to optimize and improve it based on real-time user feedback, making it unable to adapt to rapid changes in the market environment and user needs.

[0004] With the development of big data and artificial intelligence technologies, user profiling technology has gradually emerged, which can construct detailed characteristic models of users by collecting and analyzing multi-dimensional user data. However, most existing methods for generating advertising content based on user profiles simply use some information from the user profile as a reference for advertising creation, without deeply exploring the core essence and intrinsic connections of user needs, and thus failing to generate truly customized advertising content that deeply matches user needs. Summary of the Invention

[0005] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide a method for customizing and generating advertising content based on user profiles, the method comprising:

[0006] Based on the behavioral preference trajectories and potential demand tendencies in the user profile, the core demand elements with genetic characteristics are decomposed and a demand gene sequence is generated. The demand gene sequence contains multiple evolvable demand segments and the association rules between demand segments.

[0007] Based on the evolutionary possibilities of each demand segment in the demand gene sequence, an AI algorithm is invoked to generate multiple sets of basic variants of advertising content with differentiated characteristics. Each basic variant of advertising content carries a gene identifier corresponding to the demand segment.

[0008] The basic variants of the advertising content are subjected to evolutionary iteration processing. Based on the association rules between demand segments in the demand gene sequence, the text structure, visual form and information hierarchy of the basic variants of the advertising content are adjusted to generate multiple generations of advertising content evolution variants, forming an advertising content evolution variant evolution sequence.

[0009] Collect user interaction feedback information corresponding to the evolutionary sequence of the advertising content evolution variants, identify the advertising content evolution variant with the highest fit with the demand gene sequence based on the user interaction feedback information, and determine the optimal evolutionary path;

[0010] The advertising content evolution variants are optimized along the optimal evolution path, and the core adaptation features in the advertising content evolution variants are integrated to generate a set of customized advertising content that is deeply aligned with the user profile's needs.

[0011] In another aspect, embodiments of the present invention also provide an advertising content customization generation system based on user profiles, including a processor and a machine-readable storage medium connected to the processor. The machine-readable storage medium is used to store programs, instructions, or code, and the processor is used to execute the programs, instructions, or code in the machine-readable storage medium to implement the above-described method.

[0012] Based on the above, this invention, by deconstructing core demand elements with genetic characteristics from user profiles based on behavioral preference trajectories and potential demand tendencies, generates demand gene sequences. This allows for in-depth exploration of the essential characteristics and intrinsic connections of user needs. Based on the evolutionary possibilities of each demand segment in the demand gene sequence, AI algorithms are used to generate multiple sets of basic variants of advertising content with differentiated characteristics, each carrying a gene identifier corresponding to a demand segment. This achieves diversified generation of advertising content, satisfying the personalized needs of different user groups. The basic variants of advertising content undergo evolutionary iteration processing, adjusting the text structure, visual form, and information hierarchy according to the association rules between demand segments. This generates multiple generations of evolving advertising content variants and forms an evolutionary sequence, enabling advertising content to dynamically evolve with changes in user needs, enhancing the adaptability and timeliness of advertising. By collecting user interaction feedback information and identifying the advertising content evolution variants with the highest compatibility with the demand gene sequence, the optimal evolution path can be determined. This allows for timely understanding of users' true reactions to advertising content. The advertising content evolution variants can be optimized along the optimal evolution path and core adaptation features can be integrated to generate a set of customized advertising content that deeply matches the user profile's demand genes. This greatly improves the accuracy and conversion rate of advertising, effectively enhancing the effectiveness of advertising and the marketing benefits for enterprises. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the execution flow of the user profile-based advertising content customization generation method provided in the embodiments of the present invention.

[0014] Figure 2 This is a schematic diagram of exemplary hardware and software components of the user profile-based advertising content customization generation system provided in this embodiment of the invention. Detailed Implementation

[0015] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a flowchart illustrating a user profile-based advertising content customization generation method according to an embodiment of the present invention. The following is a detailed description of this user profile-based advertising content customization generation method.

[0016] Step S110: Based on the behavioral preference trajectory and potential demand tendencies in the user profile, decompose the core demand elements with genetic characteristics and generate a demand gene sequence. The demand gene sequence contains multiple evolvable demand segments and the association rules between demand segments.

[0017] In this embodiment, taking a user in the smart home appliance field as an example, the user profile encompasses various behavioral data and demand-related information of the user on the smart home appliance platform. Behavioral preference trajectories reflect the user's attention to and selection of smart home appliance products at different times, while potential demand tendencies reflect the user's unexpressed but potentially existing expectations regarding the functions and scenarios of smart home appliances. Through in-depth analysis of this information, core demand elements with stability and continuity can be identified and combined into a demand gene sequence. Each demand segment in the demand gene sequence represents a specific demand of the user in a certain aspect, while association rules illustrate the mutual influence and evolutionary relationships between different demand segments.

[0018] Step S111: Extract behavioral preference trajectories from user profiles. The behavioral preference trajectories include the user's content browsing trajectory, product selection trajectory, and service usage trajectory in different periods. Each trajectory carries a time sequence marker and a description of the intensity of the behavior.

[0019] In the smart home appliance scenario, content browsing history can be a record of users viewing product introduction pages for different categories such as smart refrigerators, smart washing machines, and smart air conditioners on the platform. For example, a user might browse the energy-saving function introduction page for smart refrigerators every Monday evening in the first quarter, and view content related to the drying function of smart washing machines every Wednesday afternoon in the second quarter. Product selection history reflects users' behavior of adding specific brands and models of smart home appliances to their favorites, comparison lists, or shopping carts. For example, a user might add a 500-liter capacity smart refrigerator from brand A to their favorites in March, and add a front-loading smart washing machine from brand B to their shopping cart in April. Service usage history includes users' after-sales service appointments and online inquiries for purchased smart home appliances. For example, a user might schedule installation service one month after purchasing a smart air conditioner, and inquire about filter replacement three months later. The time sequence of each history is marked precisely to the year, month, day, hour, and minute. The intensity of the behavior is quantified by the user's dwell time and frequency of operation on relevant pages. For example, a user who spends 20 minutes on the energy-saving function page of a smart refrigerator and adds it to their favorites three times indicates a relatively high intensity of behavior.

[0020] Step S112: Extract potential demand tendencies from user profiles. These potential demand tendencies include descriptions of users' extended needs for existing content, descriptions of their needs for exploring unexplored areas, and descriptions of their needs for upgrading known services.

[0021] For smart home appliance users, extended needs descriptions might include requests for features beyond existing smart refrigerators, such as a user commenting, "It would be great if my refrigerator could remind me which foods are about to expire." Exploration needs in unfamiliar areas could stem from questions users leave while browsing smart security product descriptions, such as, "How do I operate the remote control function of a smart door lock, and how secure is it?" This indicates a user's willingness to explore this unexperienced area. Upgrade needs for known services might be reflected in user feedback on smart home appliance after-sales service, such as, "I hope the after-sales service response time can be faster, ideally resolving issues within 24 hours," reflecting a need for upgrades to existing after-sales services.

[0022] Step S113: The behavioral preference trajectory is fragmented and key behavioral segments are selected from high to low according to the behavioral intensity description. Each key behavioral segment corresponds to a potential demand point.

[0023] To analyze the behavioral preferences of smart home appliance users, the behavior is broken down into segments based on time periods and product categories. For example, user behavior in the first quarter is broken down into segments such as browsing smart refrigerators and smart washing machines. These segments are then ranked according to their intensity, and segments with higher intensity are selected as key behavioral segments. For instance, if a user spends the most time and operates most frequently on the smart refrigerator's energy-saving function page, then that browsing segment is a key behavioral segment, indicating a high level of user interest in the energy-saving function. Similarly, if a user adds a smart washing machine to their shopping cart and views it multiple times, that product's selected behavior segment is a key behavioral segment, indicating a potential purchase intention for that smart washing machine.

[0024] Step S114: Extract the elements of the potential demand tendency and extract the core demand attributes, which include the functional orientation, scenario orientation and experience orientation of the demand.

[0025] This analysis examines the potential needs of smart home appliance users to extract core demand attributes. In terms of functionality, the desire for a food expiration reminder function in a smart refrigerator reveals the need for "food management and reminders." Regarding scenarios, considering family size and lifestyle habits, a user's potential need for a smart oven might include a scenario related to "cooking for large gatherings." Finally, the focus shifts to user expectations regarding after-sales service response speed, resulting in the extraction of the "efficient and convenient after-sales service experience" as the desired outcome.

[0026] Step S115: Bind key behavioral fragments with core requirement attributes to form core requirement elements. Each core requirement element contains the basis for the behavior and the direction of the requirement.

[0027] By linking the key behavioral segment of browsing the energy-saving function of a smart refrigerator with the core demand attribute of "energy-saving function," a core demand element can be formed. The behavioral basis is the high dwell time and multiple operation records of users on the smart refrigerator's energy-saving function page, indicating a demand for the energy-saving performance of smart refrigerators. Similarly, by linking the key behavioral segment of purchasing intention for a smart washing machine with the core demand attribute of "large-capacity drum washing," the behavioral basis is the user's record of adding the washing machine to their shopping cart and viewing it multiple times, indicating a demand for large-capacity drum washing machines.

[0028] Step S116: Based on the time sequence of behavioral preference trajectories, sort the core demand elements and define the evolutionary association rules between adjacent core demand elements. Integrate the sorted core demand elements and evolutionary association rules to generate a demand gene sequence. The evolutionary association rules are used to describe the process by which the previous core demand element guides the change of the next core demand element.

[0029] Based on chronological order, assuming a user first focuses on the energy-saving function of a smart refrigerator and then on the drying function of a smart washing machine, the corresponding core demand elements are ranked as follows: "energy-saving demand for smart refrigerators" first, followed by "drying demand for smart washing machines." The evolutionary relationship between these two can be defined as follows: the user's focus on the energy-saving function of smart appliances gradually extends to the focus on functions that improve convenience (such as drying). Integrating multiple such ordered core demand elements and their corresponding evolutionary relationship rules generates a demand gene sequence. For example, the demand gene sequence might sequentially include energy-saving demand for smart refrigerators, drying demand for smart washing machines, and temperature control demand for smart air conditioners, with corresponding evolutionary relationship rules between each adjacent core demand element.

[0030] Step S120: Based on the evolutionary probability of each demand segment in the demand gene sequence, call the AI ​​algorithm to generate multiple sets of basic variants of advertising content with differentiated characteristics. Each basic variant of advertising content carries a gene identifier corresponding to the demand segment.

[0031] In the context of smart home appliances, each demand segment in the demand gene sequence has multiple evolutionary directions. For example, the demand segment "energy-saving demand for smart refrigerators" may evolve into directions such as detailed introductions to the principles of energy-saving technologies and comparisons of different energy-saving modes. By calling a trained AI algorithm, multiple sets of basic variations of advertising content can be generated based on the characteristics and evolutionary possibilities of the demand segments. These variations differ in text description, visual presentation, and information organization, but are all associated with the corresponding demand segments, and each variation carries a unique genetic identifier for differentiation and traceability.

[0032] Step S121: Analyze each demand segment in the demand gene sequence, extract the core demand attributes and evolutionary association rules of each demand segment, and determine the content direction that each demand segment can extend to.

[0033] Analyzing the "smart washing machine drying demand" segment in the demand gene sequence, the core demand attributes are "efficient drying and garment care." The evolutionary association rules may shift from a focus on drying efficiency to a focus on drying modes for different fabrics. Based on this, potential extension areas include the working principles of drying technology, drying program settings for different fabrics, and the balance between drying effect and energy saving. For the "smart air conditioner temperature control demand" segment, the core demand attributes are "precise temperature control and quiet operation." The evolutionary association rules may shift from basic temperature control needs to a focus on sleep mode temperature control. Potential extension areas include the accuracy of smart temperature sensors, the temperature adjustment logic in sleep mode, and the implementation methods of quiet operation technology.

[0034] Step S122: Input the core requirement attributes of each requirement fragment into the AI ​​algorithm, generate a draft of text content based on the functional orientation of the requirement attributes, generate a draft of visual style based on the scenario orientation of the requirement attributes, and generate a draft of information hierarchy based on the experience orientation of the requirement attributes.

[0035] Inputting the core requirement of "energy-saving smart refrigerator needs" into the AI ​​algorithm, with the function directed towards "energy saving," the initial draft of the text content generated by the AI ​​algorithm might include descriptions such as "This smart refrigerator uses a new type of inverter compressor, effectively reducing energy consumption, with daily power consumption as low as a certain level." If the scenario is "daily family use," the initial visual style might use a warm family kitchen background, with the smart refrigerator placed within it, highlighting the energy-saving label. If the user experience is "easy access to energy-saving data," the initial information hierarchy would place information such as the refrigerator's daily power consumption data and energy-saving mode switching methods in a prominent position for easy access by the user.

[0036] Step S123: Make differentiated adjustments to the initial draft of the text content by changing the sentence structure, replacing core descriptive words, and adjusting the tone and style to generate multiple sets of advertising content text variations. Each set of advertising content text variations maintains consistency with the core requirement attributes of the requirement fragment.

[0037] The initial draft of the text content regarding the "energy-saving needs of smart refrigerators" was revised by changing the sentence structure. The phrase "This smart refrigerator uses a new type of inverter compressor, effectively reducing energy consumption, with daily power consumption as low as a certain level" was changed to "The new inverter compressor is applied to this smart refrigerator, significantly reducing energy consumption, with daily power consumption only at a certain level." Key descriptive terms were replaced, such as "reducing energy consumption" or "saving electricity costs." The tone was also adjusted from an objective statement to a more approachable one, such as "Saving every kilowatt-hour for your family, this smart refrigerator uses a new type of inverter compressor, with daily power consumption as low as a certain level!" Multiple variations of the advertising text were generated using these methods, each focusing on the core requirement of energy saving.

[0038] Step S124: Make differentiated adjustments to the initial visual style draft by changing the color scheme, adjusting the shape of graphic elements, and transforming the layout structure to generate multiple sets of visual variations of the advertising content. Each set of visual variations of the advertising content meets the basic requirements of the scene to which the demand segment is directed.

[0039] For the initial draft of the visual style for "Smart Refrigerator Energy Saving Needs," the color scheme was changed, shifting the primary color from blue to green, as green is often associated with energy saving and environmental protection. The graphic elements were adjusted, changing the energy-saving logo from a circle to a square, and refining the compressor graphic. The layout was also changed, moving the refrigerator from the left to the right side of the image, with the energy-saving data charts positioned accordingly. All generated visual variations of the advertisement are based on everyday household usage scenarios, aligning with the requirements of scenario-based advertising.

[0040] Step S125: Make differentiated adjustments to the initial draft of the information hierarchy by changing the information display order, adjusting the weight of information modules, and changing the information delivery rhythm to generate multiple sets of advertising content information variations. Each set of advertising content information variations meets the core standards of the required segment experience.

[0041] Regarding the initial draft of the information hierarchy for "smart refrigerator energy-saving needs," the information display order was changed, placing information on energy-saving mode switching methods before daily power consumption data. The weight of information modules was adjusted, increasing the length of the module explaining the principles of energy-saving technology while reducing the proportion of other non-core functional modules. The information delivery rhythm was changed, splitting the originally continuous information into multiple sections with brief pauses between each section. The generated variations of advertising content all allow users to easily understand energy-saving data, meeting the core standards for user experience.

[0042] Step S126: Combine each set of advertising content text variants, advertising content visual variants, and advertising content information variants to form multiple sets of advertising content basic variants. Add a gene identifier that matches the corresponding demand fragment to each advertising content basic variant. The gene identifier includes the demand fragment number and the core demand attribute code.

[0043] A set of text variations, a set of visual variations, and a set of information variations related to the "energy-saving needs of smart refrigerators" are combined to form a basic set of advertising content variations. A genetic identifier is added to this variation; assuming the need fragment number is XQ001 and the core need attribute code is JN001, then the genetic identifier is XQ001-JN001. Similarly, other combinations forming basic advertising content variations also have corresponding genetic identifiers added, such as XQ001-JN002, XQ001-JN003, etc., for subsequent evolution and tracking.

[0044] Step S130: Perform evolutionary iteration processing on the basic variant of the advertising content. Based on the association rules between demand segments in the demand gene sequence, adjust the text structure, visual form and information hierarchy of the basic variant of the advertising content to generate multiple generations of advertising content evolution variants, forming an advertising content evolution variant evolution sequence.

[0045] In the generation of smart home appliance advertisements, based on the association rules from "energy-saving demand for smart refrigerators" to "drying demand for smart washing machines" in the demand gene sequence, the basic variant of the advertisement content corresponding to the "energy-saving demand for smart refrigerators" is evolved and iterated. Its text structure is adjusted to add connecting content related to the drying function of smart washing machines; the visual form is changed to incorporate visual elements of washing machines; and the information hierarchy is adjusted to reserve display space for subsequent drying demand information. After the above adjustments, a second-generation advertisement content evolution variant is generated. Then, based on the second-generation variant, adjustments are made according to the next association rule to generate the third generation, and so on, forming an evolutionary sequence of advertisement content variants arranged in generational order.

[0046] Step S131: Extract the evolutionary association rules between adjacent demand segments in the demand gene sequence, and determine the evolutionary information of the transition from the basic variant of advertising content corresponding to the previous demand segment to the basic variant of advertising content corresponding to the next demand segment.

[0047] The evolutionary association rules from "smart washing machine drying needs" to "smart air conditioner temperature control needs" in the demand gene sequence were extracted. These rules likely reflect users' demand for smart home appliances to enhance convenience, extending from washing and drying to indoor environmental comfort control. Based on this, the evolutionary information determined is that advertising content should reflect the natural transition from the convenience of clothes drying to comfortable indoor temperature control, emphasizing the commonality between the two in improving quality of life.

[0048] Step S132: Starting with the first-generation basic variant of advertising content, adjust the text structure of the previous basic variant of advertising content according to the guiding relationship in the evolutionary association rules. While retaining the core information of the previous basic variant of advertising content text, add text description elements corresponding to the next demand fragment to form the second-generation advertising content text evolution variant.

[0049] The first-generation advertising content variant was designed for the "smart washing machine drying needs," with its core text revolving around efficient drying and garment care. Based on the guiding relationships of evolutionary association rules, while retaining these core messages, textual descriptive elements related to the "smart air conditioner temperature control needs" were added, such as, "After enjoying the convenience of washing and drying clothes instantly, a comfortable indoor temperature is equally important. Our brand's smart air conditioner will bring you a precise temperature control experience." These adjustments resulted in the second-generation advertising content text evolution variant.

[0050] Step S133: Adjust the visual form of the previous basic variant of the advertising content. While maintaining the visual style consistency of the previous basic variant of the advertising content, incorporate the visual element features corresponding to the next demand segment to form the second-generation visual evolution variant of the advertising content.

[0051] The visual style of the previous basic variant of the advertisement (related to smart washing machines) was a warm family scene with a predominantly warm color scheme. When adjusting the visual form, the unity of the warm color scheme and the family scene was maintained, and visual elements of smart air conditioners were incorporated, such as adding an image of the smart air conditioner in the corner of the image, or subtly showing a diagram of the air conditioner vent in the background, thus forming the second-generation visual evolution variant of the advertisement.

[0052] Step S134: Adjust the information hierarchy of the previous basic variant of the advertising content. While maintaining the information transmission logic of the previous basic variant of the advertising content, increase the display weight of the information module corresponding to the next demand segment, and adjust the arrangement order of the information modules to guide users to pay attention to the content of the next demand segment, thus forming the second-generation advertising content information evolution variant.

[0053] The information delivery logic of the previous basic variant of the advertisement was to first introduce the advantages of the drying function of the smart washing machine, and then explain how to use it. When adjusting the information hierarchy, this logic is maintained, while the display weight of the information module corresponding to the temperature control needs of the smart air conditioner is increased. It is moved from its original secondary position to a prominent position near the end, and the arrangement order guides users to naturally transition to the air conditioner information after learning about the washing machine information, forming the second-generation variant of the advertisement content information evolution.

[0054] Step S135: Combine the second-generation advertising content text evolution variant, the second-generation advertising content visual evolution variant, and the second-generation advertising content information evolution variant to form the second-generation advertising content evolution variant, and add an evolutionary generation identifier and corresponding demand fragment association code to the second-generation advertising content evolution variant.

[0055] The text evolution variant formed in step S132, the visual evolution variant formed in step S133, and the information evolution variant formed in step S134 are combined to form a complete second-generation advertising content evolution variant. An evolutionary generation identifier "D2" and corresponding demand fragment association codes "XQ002-XQ003" are added to it, where XQ002 represents the demand fragment for intelligent washing machine drying, and XQ003 represents the demand fragment for intelligent air conditioner temperature control.

[0056] Step S136: Repeat the above adjustment steps to generate a third-generation advertising content evolution variant based on the second-generation advertising content evolution variant, and so on, until a multi-generation advertising content evolution variant covering all demand segments of the demand gene sequence is generated, and arranged in generational order to form an advertising content evolution variant evolution sequence.

[0057] Based on the second-generation advertising content evolution variants, and according to the evolutionary association rules from "intelligent air conditioning temperature control demand" to the next demand segment (such as intelligent security demand), the adjustment process of steps S132 to S135 is repeated to generate the third-generation advertising content evolution variants. This process continues until the generated advertising content evolution variants cover all demand segments in the demand gene sequence. These multi-generation advertising content evolution variants are then arranged in generational order "D1, D2, D3..." to form an advertising content evolution variant evolution sequence.

[0058] Step S1321: Extract the core information statements from the basic variant text structure of the previous advertisement content. The core information statements include the functional description, user benefit description, and scenario application description of the previous requirement fragment.

[0059] Taking a basic variation of the previous advertisement (related to the drying needs of a smart washing machine) as an example, we can extract its core information statements. The function description statement is "This smart washing machine is equipped with heat pump drying technology, which enables clothes to be dried quickly without damaging the fabric"; the user benefit statement is "Users do not need to wait for clothes to air dry naturally, saving time and space, and clothes always remain soft and fluffy"; the scenario application description statement is "Suitable for various scenarios such as daily household laundry and drying clothes in rainy weather".

[0060] Step S1322: Analyze the guiding relationship in the evolutionary association rules to determine the association nodes between the core information of the previous requirement segment and the content of the next requirement segment. The association nodes are functional attributes or scene attributes shared by the previous requirement segment and the next requirement segment.

[0061] This analysis examines the evolutionary relationship between "smart washing machine drying needs" and "smart air conditioner temperature control needs," highlighting that both aim to improve the convenience and comfort of family life. The key link is the shared scenario attribute of "improved quality of family life," as both drying clothes and controlling indoor temperature ultimately contribute to enhancing the overall quality of family life. Furthermore, both possess the functional attribute of "intelligent adjustment"—smart washing machines can intelligently adjust drying parameters based on the fabric type, and smart air conditioners can intelligently adjust temperature based on the indoor environment—this functional attribute also constitutes a key link.

[0062] Step S1323: Using the associated node as the connection point, add a transition statement after the core information statement of the previous advertisement content variant. The transition statement should clearly define the connection relationship between the associated node and the content of the next requirement segment.

[0063] Using the aforementioned related nodes "improving the quality of family life" and "intelligent adjustment" as connecting points, a transitional statement is added after the core information statement of the previous basic variation of the advertisement content (related to the drying needs of smart washing machines). For example, "Just as smart washing machines improve the convenience of family life and create a high-quality laundry experience through intelligent drying, intelligent adjustment technology also plays an important role in controlling the comfort of the home environment." This transitional statement clarifies the connection between the intelligent drying function of smart washing machines and the intelligent temperature control function of smart air conditioners in terms of "improving the quality of family life" and "intelligent adjustment," achieving a natural transition from washing machine needs to air conditioner needs.

[0064] Step S1324: Extract the text description element corresponding to the next requirement fragment. The text description element includes the functional extension description, new scenario application description and added value introduction of the next requirement fragment.

[0065] The next requirement segment is "Smart Air Conditioner Temperature Control Needs," and its text description elements are extracted. The extended function description is: "This smart air conditioner uses dual-frequency temperature control technology, which can automatically adjust the temperature according to the number of people in the room and the light intensity, with a temperature control accuracy of ±0.5℃"; the new scenario application description is: "Suitable for different family activity scenarios such as bedroom sleeping scenarios and living room gathering scenarios, ensuring comfortable temperature in each scenario"; the added value description is: "Equipped with silent operation technology, the operating noise is as low as 20 decibels, and it also has an air purification function to improve indoor air quality." These text description elements comprehensively cover the core content of the smart air conditioner temperature control needs.

[0066] Step S1325: Integrate the transition statement with the text description element of the next requirement fragment and add it to the text structure of the previous advertisement content base variant.

[0067] Integrate the transitional statement formed in step S1323 with the text description elements extracted in step S1324 to form a complete text content, such as: "Just as smart washing machines improve the convenience of family life and create a high-quality laundry experience through intelligent drying, intelligent adjustment technology also plays an important role in the comfortable control of the home environment. This smart air conditioner adopts dual-frequency temperature control technology, which can automatically adjust the temperature according to the number of people in the room and the light intensity, with a temperature control accuracy of ±0.5℃; it is suitable for different family activity scenarios such as bedroom sleeping scenarios and living room gathering scenarios, ensuring temperature comfort in each scenario; it is equipped with silent operation technology, with operating noise as low as 20 decibels, and also has an air purification function to improve indoor air quality." Add this integrated text to the end of the text structure of the previous advertisement content basic variant to expand the content.

[0068] Step S1326: Redundant information is removed from the basic variant text structure of the adjusted previous advertisement content. Conflicting or repetitive expressions with the content of the next requirement segment are deleted. The integrity of the core information of the previous requirement segment and the newly added descriptions of the next requirement segment are retained to form the second-generation advertisement content text evolution variant.

[0069] Examine the adjusted text structure. If the text of the previous basic variant of the advertisement contains the statement "laundry is a major household chore in family life," while the newly added text description of the smart air conditioner emphasizes the improvement of comfort in many aspects of family life, there is no conflict in the wording. However, there may be some repetitive descriptions related to "family life," so duplicate words can be appropriately simplified. If the previous text mentions "no need to worry about the weather affecting clothes drying," and the new text has no related conflicting content, then no deletion is necessary. After removing redundant information, ensure that the core information of the previous demand segment (smart washing machine drying function, user benefits, application scenarios) is fully retained, and the new description of the subsequent demand segment (smart air conditioner temperature control function, scenarios, added value) is clearly presented, ultimately forming the second-generation variant of the advertisement content text evolution.

[0070] Step S140: Collect user interaction feedback information corresponding to the evolutionary sequence of the advertising content evolution variant, identify the advertising content evolution variant with the highest fitness to the demand gene sequence based on the user interaction feedback information, and determine the optimal evolutionary path.

[0071] During the smart home appliance advertising display process, the evolutionary sequence of advertising content variants is pushed to users in generational order, "D1, D2, D3...". All user interactions during browsing are recorded, forming user interaction feedback information. By analyzing this information, it is determined which generation of advertising content variant best matches the user's needs and gene sequence, thereby identifying the evolutionary path from the basic variant to this optimal variant.

[0072] Step S141: Display the different generations of advertising content evolution variants in the evolutionary sequence of the advertising content evolution variants in generational order, and collect user interaction feedback information for each advertising content evolution variant. The user interaction feedback information includes the user's browsing time record, content click record, and feedback record for the advertising content evolution variant. Each record carries the generational identifier and gene identifier of the corresponding advertising content evolution variant.

[0073] The system sequentially displays different versions of advertising content to users, following a generational order: D1 (related to energy-saving needs in smart refrigerators), D2 (related to the transition from drying needs in smart washing machines to temperature control needs in smart air conditioners), and D3 (related to the transition from temperature control needs in smart air conditioners to smart security needs). When a user views a D2 generation advertisement, the system records that they viewed the advertisement for 1 minute and 30 seconds, clicked the link introducing the temperature control accuracy of smart air conditioners in the advertisement, and left a message in the feedback box stating "very interested in the quiet function of air conditioners." These records are labeled with the corresponding generational identifier "D2" and gene identifier "XQ002-XQ003" for subsequent analysis.

[0074] Step S142: Classify and statistically analyze user interaction feedback information. Statistically analyze the average browsing time, core content click-through rate, and positive feedback percentage of each generation of advertising content evolution variants according to generational identifiers. Statistically analyze the interaction data of advertising content evolution variants corresponding to the demand segment according to gene identifiers.

[0075] According to generational identifiers, the average viewing time for D1 generation ads was 50 seconds, with a 15% click-through rate (CTR) for the core content (introduction to the energy-saving function of smart refrigerators) and a 30% positive feedback rate. For D2 generation ads, the average viewing time was 1 minute and 20 seconds, with a 25% CTR for the core content (introduction to the temperature control function of smart air conditioners) and a 60% positive feedback rate. For D3 generation ads, the average viewing time was 45 seconds, with a 10% CTR for the core content (introduction to smart security functions) and a 25% positive feedback rate. According to gene identifiers, for XQ001 (energy-saving needs of smart refrigerators), the average viewing time for ads was 60 seconds, with a CTR of 18%. For XQ002-XQ003 (transition from smart washing machine drying to air conditioner temperature control), the average viewing time was 1 minute and 15 seconds, with a CTR of 22%.

[0076] Step S143: Analyze the relationship between the interaction data and the demand gene sequence of each generation of advertising content evolution variants. When the average browsing time of any generation of advertising content evolution variants is higher than that of other generations of advertising content evolution variants, the core demand attribute of the core content click rate corresponds to the core demand segment, and the positive feedback ratio includes the recognition description of demand extension, mark that generation as a candidate suitable generation.

[0077] Analyzing the interaction data across generations, the average viewing time for D2 generation ads (1 minute 20 seconds) is higher than the 50 seconds for D1 generation and the 45 seconds for D3 generation. The 25% click-through rate for D2 generation's core content corresponds to the core demand attributes of smart air conditioner temperature control: "precise temperature control and quiet operation," which matches the core attributes of the demand segment. Among the 60% positive feedback from D2 generation, there are descriptions of recognition of extended demand, such as "the quiet function of the air conditioner is very practical" and "the high temperature control accuracy is very suitable for bedroom use." Based on these factors, D2 generation is marked as a candidate suitable generation.

[0078] Step S144: Extract the core features of each advertising content evolution variant in the candidate adaptation generation. The core features include the description of related nodes in the text structure of the advertising content evolution variant, the element fusion method in the visual form of the advertising content evolution variant, and the weight allocation logic in the information hierarchy of the advertising content evolution variant.

[0079] Core features were extracted from the evolutionary variations of D2 generation advertising content. The related nodes in the text structure are described as follows: "Just as smart washing machines enhance the convenience of family life and create a high-quality laundry experience through intelligent drying, intelligent adjustment technology also plays an important role in controlling the comfort of the home environment." In terms of visual form, the element integration method is to use a smart washing machine as the main display object in a warm family scene, with a thumbnail of a smart air conditioner embedded in the upper right corner of the image, maintaining a unified warm color scheme. The weighting logic in the information hierarchy is to first introduce the drying function of the smart washing machine in detail (60%), then use transitional content to introduce the temperature control function of the smart air conditioner (40%), highlighting the information modules on temperature control accuracy and quiet operation.

[0080] Step S145: Compare the degree of fit between the core features of each advertising content evolution variant in the candidate adaptation generation and the demand gene sequence, and select the advertising content evolution variant with the best interaction data that completely matches the association rules of the core features and demand fragments.

[0081] Comparing the core features and demand gene sequence of D2 generation ads, the association rule from "smart washing machine drying demand" to "smart air conditioner temperature control demand" in the demand gene sequence reflects the extension of users' needs for convenience from washing to environmental control. The description of the association nodes in the D2 generation text embodies this extension relationship, the visual element integration method meets the requirements of scene uniformity, and the information hierarchy weight allocation logic also aligns with the focus of the demand transition. The core features and association rules are perfectly matched. Furthermore, D2 generation has the best average browsing time, core content click-through rate, and positive feedback ratio among all generations. Therefore, the D2 generation ad content evolution variant is selected as the optimal adapted ad content evolution variant.

[0082] Step S146: Trace the generation process of the optimal adapted advertising content evolution variant, determine the evolution adjustment steps of the optimal adapted advertising content evolution variant from the basic advertising content variant to the current generation, sort out the adjustment logic in generational order, and form the optimal evolution path. The optimal evolution path includes the adjustment direction, related nodes and core feature retention strategy of each generation of advertising content evolution variant.

[0083] Tracing the generation process of D2 generation ads, it evolved from the basic variant of D1 generation ad content (related to the drying needs of smart washing machines). The adjustment steps included: adding transitional statements and descriptive elements related to smart air conditioner temperature control to the text structure while retaining core information about smart washing machines; incorporating visual elements of smart air conditioners to maintain a warm color scheme for the home scene; and increasing the weight of the air conditioner temperature control information module in the information hierarchy. Following the generational order of the adjustment logic, the optimal evolutionary path is as follows: starting with the basic variant of smart washing machine drying needs (D1), through textual association nodes, visual element integration, and information weight adjustments, a D2 generation evolutionary variant that meets the extended needs is generated. This path clarifies that the adjustment direction of each generation is to transition towards the needs of environmental comfort control, with the association nodes being "improvement of family life quality" and "intelligent adjustment," and the core feature retention strategy being to retain the core functional descriptions and scene style of the previous need.

[0084] Step S1431: Calculate the average browsing time of all generational advertising content evolution variants, determine the baseline value of the average browsing time, and when the average browsing time of a certain generation of advertising content evolution variants exceeds the baseline value by a certain percentage, include the generation corresponding to that generation of advertising content evolution variants in the preliminary screening of generations.

[0085] The total average browsing time for Generation D1 (50 seconds), Generation D2 (1 minute 20 seconds, 80 seconds), and Generation D3 (45 seconds) is calculated to be 50 + 80 + 45 = 175 seconds. The total number of generations is 3, and the baseline average browsing time is 175 ÷ 3 ≈ 58.33 seconds. A standard is set at 20% exceeding this baseline. 20% of the baseline is 58.33 × 20% ≈ 11.67 seconds, and the baseline plus 20% is 58.33 + 11.67 = 70 seconds. Generation D2's 80 seconds exceeds 70 seconds; therefore, Generation D2 is included in the initial generation selection.

[0086] Step S1432: Analyze the click-through rate of the core content of the initially screened generational advertising content evolution variants, extract the core demand attributes of the demand fragment corresponding to the core content of each advertising content evolution variant, and count the proportion of click behavior concentrated in the description area of ​​the core demand attribute. When the proportion exceeds the preset proportion, it is determined that the click-through rate of the generational advertising content evolution variant matches the core demand attribute of the demand fragment.

[0087] The initial selection of the advertising generation is D2, whose core content is an introduction to the temperature control function of intelligent air conditioners. The corresponding core demand attribute of the demand segment is "precise temperature control and quiet operation". User click behavior in the D2 generation ads was analyzed. The total number of clicks was 100, with 40 clicks on the "temperature control precision introduction" area, 35 clicks on the "quiet operation technology" area, and 25 clicks on other non-core areas. The number of clicks on the core demand attribute description area was 40 + 35 = 75, accounting for 75 ÷ 100 = 75% of the total clicks. The preset percentage is 60%, and 75% exceeds the preset percentage. Therefore, the click-through rate of the D2 generation ads is determined to match the core demand attribute of the demand segment.

[0088] Step S1433: Collect positive feedback on the advertising content evolution variants of the preliminary screening generation, extract the number of feedback opinions containing descriptions of demand extension recognition, calculate the proportion of this number to the total feedback opinions, and when the proportion exceeds a preset threshold, determine that the advertising content evolution variant of that generation has obtained demand extension recognition.

[0089] A total of 50 positive feedback comments were collected regarding the D2 generation advertisement. Among them, 35 comments included descriptions of extended demand recognition, such as "The extension from washing machine drying to air conditioning temperature control perfectly meets family needs" and "It not only solves laundry problems but also considers indoor comfort, which is very thoughtful." This number accounts for 35 ÷ 50 = 70% of the total feedback comments. The preset threshold is 50%, and 70% exceeds the preset threshold; therefore, it is determined that the D2 generation advertisement content evolution variant has achieved extended demand recognition.

[0090] Step S1434: Generations that simultaneously meet the following criteria are marked as candidate compatible generations: average browsing time duration, core content click-through rate matching requirement fragment core requirement attributes, and positive feedback including requirement extension recognition description.

[0091] The average viewing time of D2 generation ads is 80 seconds, which exceeds the benchmark value and proportion requirements. The click-through rate of core content is 75%, which exceeds the preset proportion and matches the core demand attributes. The proportion of extended demand recognition descriptions in positive feedback is 70%, which exceeds the preset threshold. Since all three conditions are met, D2 generation is marked as a candidate generation for adaptation.

[0092] Step S1435: Prioritize the candidate adaptation generations and sort them from high to low based on the comprehensive score of the average browsing time of the ad content evolution variants within the candidate adaptation generation, the core content click-through rate, and the proportion of demand extension recognition. The higher the comprehensive score, the higher the priority of the candidate adaptation generation.

[0093] If multiple candidate generations exist, assuming generation D4 also meets the criteria, generation D2 has an average browsing time of 80 seconds, a core content click-through rate of 75%, and a demand extension approval rate of 70%; generation D4 has an average browsing time of 75 seconds, a core content click-through rate of 70%, and a demand extension approval rate of 65%. Assuming equal weights for all three metrics, the overall score is calculated by adding (or averaging) the values ​​of each metric. Generation D2's overall score is 80 + 75 + 70 = 225, while generation D4's overall score is 75 + 70 + 65 = 210. Therefore, generation D2 has a higher priority than generation D4. In this embodiment, only generation D2 is a candidate generation, and its priority is the highest.

[0094] Step S150: Optimize the advertising content evolution variants along the optimal evolution path, integrate the core adaptation features in the advertising content evolution variants, and generate a set of customized advertising content that deeply matches the user profile's needs.

[0095] Following the optimal evolutionary path, and based on the optimal adaptation variant of D2 generation ad content, the core adaptation features are further optimized. High-quality features from text, visual, and informational dimensions are integrated to generate multiple sets of customized ad content that meet user needs, forming a collection for subsequent display.

[0096] Step S151: Analyze the adjustment steps of each generation of advertising content evolution variants in the optimal evolution path, and extract the core adaptation features retained in each generation of advertising content evolution variants. The core adaptation features include the related node descriptions at the text level of the advertising content evolution variants, the element fusion style at the visual level of the advertising content evolution variants, and the weight allocation mode at the information level of the advertising content evolution variants.

[0097] This analysis examines the adjustment steps from generation D1 to generation D2 in the optimal evolution path, extracting the core adaptation features retained in generation D1: at the textual level, the statement "new variable frequency compressors achieve energy saving"; at the visual level, the fusion style of a warm family kitchen scene with energy-saving labels; and at the information level, the weighted distribution pattern prioritizing the display of energy-saving data. The core adaptation features retained in generation D2 are extracted: at the textual level, the related node statement "just as smart washing machines improve the convenience of family life through intelligent drying..."; at the visual level, the warm color fusion style of washing machine and air conditioner elements in the family scene; and at the information level, the weighted distribution pattern of 60% for washing machine content and 40% for air conditioner content.

[0098] Step S152: Classify and integrate the extracted core adaptation features, and establish feature sets according to the three dimensions of text, vision and information. Each feature set contains the core adaptation features of each generation of advertising content evolution variants in the optimal evolution path.

[0099] The text-dimensional feature set includes: descriptions of energy-saving functions in D1 generation, descriptions of related nodes in D2 generation, and the tone and style of smart home appliance function descriptions. The visual-dimensional feature set includes: background style of home scenes, warm color schemes, and how product elements are integrated into the scene (e.g., the layout of a washing machine in the kitchen and an air conditioner in a corner of the living room). The information-dimensional feature set includes: the logic of prioritizing the display of core functional information, the proportion of content for different needs, and the arrangement order of information modules.

[0100] Step S153: Analyze the feature correlations in the feature sets of each dimension, and identify the style matching relationship between text features and visual features, the display synergy relationship between visual features and information features, and the logical support relationship between information features and text features.

[0101] Analysis revealed that the warm and friendly tone in the text dimension matches the warm color scheme and family scene integration in the visual dimension, forming a unified and warm family atmosphere; the clear display of product elements in the visual dimension works in conjunction with the priority display of core functional information in the information dimension, enabling users to quickly match visual elements with key information; the chronological arrangement of demand fragments in the information dimension provides logical support for the time transition descriptions of related nodes in the text dimension, ensuring the coherence of content delivery.

[0102] Step S154: Based on the aforementioned feature correlation, optimize and integrate the feature sets of each dimension to form a text-optimized feature group, a visual-optimized feature group, and an information-optimized feature group.

[0103] The text optimization feature group integrates a warm and friendly tone, clear descriptions of related nodes, and accurate functional descriptions to ensure logical coherence and a consistent style. The visual optimization feature group integrates warm-toned family scene backgrounds, a layout where product elements blend naturally with the scene, and logo designs that highlight core functions, creating visual harmony and emphasizing key points. The information optimization feature group integrates information modules that prioritize core functional information, allocate appropriate proportions of required content, and arrange them in a logical order, improving information delivery efficiency.

[0104] Step S155: Apply the text optimization feature group to the text creation of the advertising content to generate multiple sets of customized advertising text that meet the feature requirements. Each set of customized advertising text includes related node descriptions and extended descriptions of requirements.

[0105] In this embodiment, the text optimization feature group clarifies the creative direction. First, based on the needs and genetic sequences of smart home appliance users, it is determined that the customized advertising text needs to revolve around the core theme of improving the quality of family life, naturally connecting the drying function of the smart washing machine with the temperature control function of the smart air conditioner. In the specific creation process, a warm and friendly tone is strictly followed, avoiding the use of dry technical jargon, making it easier for users to understand and accept.

[0106] Step S1551: Extract the related node expression template and the requirement extension description framework from the text optimization feature group. The related node expression template contains the sentence structure connecting different requirement fragments, and the requirement extension description framework contains the expression logic of function extension, scenario extension, and experience extension.

[0107] The text optimization feature group extracts related node expression templates, such as the sentence structure "Just as [the previous demand product] brought [the previous benefit] to your family life through [the previous function], [the next demand product]'s [next function] will further create [the next benefit] for you." The demand extension description framework includes the following expression logic: for function extension, "Based on [the previous function], [the next function] has [new functional features]"; for scenario extension, "Whether it is [the previous applicable scenario] or [the next applicable scenario], it can meet your needs"; and for experience extension, "Not only can you enjoy [the previous experience], but you can also experience [the next experience]."

[0108] Step S1552: Determine the core demand line of the customized advertising text. The core demand line is determined based on the main demand segments of the demand gene sequence and evolutionary association rules.

[0109] Based on the demand gene sequence, the main demand segments are the demand for smart washing machine drying and smart air conditioner temperature control. The evolutionary association rule is an extension from improving the convenience of laundry to improving the comfort of the home environment. Therefore, the core demand theme of the customized advertising text is determined to be "to comprehensively improve the convenience and comfort of family life through the synergy of smart home appliance functions."

[0110] Step S1553: Based on the core demand line, select the matching related node description template, fill the core information of adjacent demand segments into the related node description template, and form the connecting paragraphs of the customized advertising text.

[0111] Select the template for the related node description: "Just as the [previous product] brings [previous benefits] to your family life through the [previous function], the [next product]'s [next function] will further create [next benefits] for you." Fill in the template with the core information of the smart washing machine's drying needs: "smart washing machine," "heat pump drying technology," and "no need to wait for clothes to dry, saving time and space," and the core information of the smart air conditioner's temperature control needs: "smart air conditioner," "dual-frequency temperature control technology," and "precise temperature control, creating a comfortable indoor environment." This will form the connecting paragraph: "Just as the smart washing machine brings the convenience of not having to wait for clothes to dry and saving time and space to your family life through heat pump drying technology, the smart air conditioner's dual-frequency temperature control technology will further create a comfortable indoor environment with precise temperature control for you."

[0112] Step S1554: Based on the requirement extension description framework, add requirement extension description content after the connecting paragraph.

[0113] Following the framework of extended description based on needs, the functional description is extended after the connecting paragraph: "Based on the heat pump drying technology that enables clothes to dry quickly without damaging the fabric, the dual-frequency temperature control technology can automatically adjust the temperature according to the number of people in the room and the light intensity, with a temperature control accuracy of ±0.5℃." The scenario extension description is: "Whether it's daily household laundry, drying clothes on rainy days, or even sleeping in the bedroom or having a gathering in the living room, it can provide you with a suitable environment." The experience extension description is: "Not only does it allow you to enjoy the comfort of clothes that always remain soft and fluffy, but it also allows you to feel a spring-like temperature in any season, enhancing the happiness of family life."

[0114] Step S1555: Adjust the overall tone and style of the customized advertising text to match the language preference features in the user profile.

[0115] User profiles indicate a preference for warm, friendly, and easily understood language. The generated text was tone-adjusted, softening the descriptions of technical parameters. For example, "Temperature control accuracy reaches ±0.5℃" was changed to "Temperature control accuracy is very high, accurate to ±0.5℃, ensuring the indoor temperature is always maintained at a comfortable level for you." More friendly expressions were added, such as appropriately using words like "you" and "our family," to enhance the emotional connection with users.

[0116] Step S1556: Perform a logical coherence check on the customized advertising text content to ensure that the related node descriptions, extended descriptions of needs, and the core need main line form a complete logical chain, and generate a set of customized advertising text.

[0117] The text content was reviewed to ensure that connecting paragraphs naturally link the functions of the smart washing machine and smart air conditioner. The extended descriptions of the requirements revolved around functions, scenarios, and experiences, all serving the core need of "comprehensively improving the convenience and comfort of family life." Logical breaks or contradictions were checked for, such as whether the extended function descriptions echoed the preceding function introductions, and whether the scenario descriptions covered the main family scenarios. After review and adjustments, a set of logically coherent customized advertising text was generated.

[0118] Step S1557: Change the description templates of different related nodes and the description framework of extended requirements, repeat the above steps, and generate multiple sets of customized advertising texts with differentiated descriptions.

[0119] The template for the related node description is changed to "When the [previous function] of the [previous product] has solved the [previous problem] for you, the [next function] of the [next product] will help you deal with the [next problem]." The requirement extension description framework emphasizes "The [next function] uses [new technology] to achieve [new effects]" in terms of function extension, and "For the [next specific scenario], the [next function] can provide [exclusive solutions]." Following steps S1553 to S1556, another set of customized advertising text is generated, creating a differentiated description from the previous text.

[0120] Step S156: Apply the visual optimization feature group to the visual design of the advertising content to generate multiple sets of customized advertising visuals that meet the feature requirements. Each set of customized advertising visuals includes elements that blend style and adapt to the required scenario.

[0121] Based on the visual optimization feature set, the visual design uses a warm-toned family scene as the background, naturally integrating product elements of the smart washing machine and smart air conditioner. For example, in a customized advertising visual, the main image is a cozy family living room, with the smart washing machine placed at the junction of the balcony and living room, and the smart air conditioner installed on the living room wall. The color scheme of both coordinates with the overall color scheme of the family scene. Additionally, an illustration of the fluffy state of clothes after drying is added next to the washing machine, and a sign indicating a comfortable indoor temperature is added next to the air conditioner, highlighting the core functions.

[0122] Step S157: Apply the information optimization feature group to the information arrangement of the advertising content to generate multiple sets of customized advertising information levels that meet the feature requirements. Each set of customized advertising information levels includes a weight allocation mode and a strategy to highlight the key needs.

[0123] In accordance with the requirements of the information optimization feature group, when customizing the hierarchical arrangement of advertising information, the core information of the smart washing machine's drying function and the smart air conditioner's temperature control function is placed in a prominent position, with a weighting of 40% and 40% respectively. The connecting information of related nodes accounts for 10%, and other auxiliary information accounts for 10%. A strategy of highlighting key needs is adopted, using bold fonts and darker colors to emphasize key information such as "heat pump drying technology," "dual-frequency temperature control technology," and "energy saving," allowing users to quickly grasp the core content.

[0124] Step S158: Combine the customized advertising text, customized advertising visuals, and customized advertising information hierarchy to form multiple complete sets of customized advertising content. Integrate all the combination schemes to generate a set of customized advertising content that deeply matches the user profile's needs.

[0125] The multiple sets of customized ad text generated in step S155, the multiple sets of customized ad visuals generated in step S156, and the multiple sets of customized ad information levels generated in step S157 are combined in different ways. For example, the first combination uses the first set of customized ad text, the first set of customized ad visuals, and the first set of customized ad information levels; the second combination uses the second set of customized ad text, the first set of customized ad visuals, and the second set of customized ad information levels, and so on. All the combination schemes are integrated to form a set of customized ad content, all of which closely align with the user profile's needs.

[0126] For example, the method may also include:

[0127] Step S210: Regularly monitor changes in the demand gene sequence of the user profile. When new key behavioral segments appear in the behavioral preference trajectory of the user profile, or new demand attributes are generated from potential demand tendencies, determine that the demand gene has mutated and generate mutated demand gene segments.

[0128] The system is designed to monitor the user profile's demand gene sequence every two weeks. During monitoring, user behavioral preference trajectories are analyzed. If it is found that users have recently been frequently browsing pages related to smart robot vacuum cleaners and spending a considerable amount of time on them, this behavior constitutes a new key behavioral segment. At the same time, if a new demand attribute of "desiring automated home cleaning" appears in the user's potential demand inclinations, it is determined that the demand gene has mutated, generating a mutated demand gene segment of "automated cleaning demand for smart robot vacuum cleaners".

[0129] Step S220: Integrate the mutated demand gene fragment into the original demand gene sequence, update the evolutionary association rules in the original demand gene sequence, and form an updated demand gene sequence.

[0130] A variant demand gene segment, "automatic cleaning demand of intelligent robotic vacuum cleaners," was added to the end of the original demand gene sequence. The evolutionary association rule of the original demand gene sequence was from the demand for drying in intelligent washing machines to the demand for temperature control in intelligent air conditioners. The updated evolutionary association rule adds "extending from the demand for temperature control in intelligent air conditioners to the demand for automated cleaning in intelligent robotic vacuum cleaners, reflecting a further upgrade in the comfort and convenience of family life," thus forming the updated demand gene sequence.

[0131] Step S230: Along the original optimal evolution path, integrate the content features corresponding to the mutant demand gene fragments into the variant advertising content of each generation, adjust the text structure of the variant advertising content of each generation to include the functional description of the mutant demand, adjust the visual form of the variant advertising content of each generation to adapt to the scene features of the mutant demand, and adjust the information hierarchy of the variant advertising content of each generation to highlight the core information of the mutant demand.

[0132] Following the original optimal evolution path, the text structure of the D1 generation ad content evolution variant is appropriately enhanced with the statement, "The convenience of family life is not only reflected in clothing handling and environmental control, but also in automated cleaning," laying the groundwork for future demand variations. In the visual form of the D2 generation ad content evolution variant, a simple graphic of a smart robotic vacuum cleaner is added to a corner of a home scene, adapting to its home cleaning scenario. Within the information hierarchy of each generation of ad content evolution variants, the core information of the smart robotic vacuum cleaner's automated cleaning—"automatic route planning and remote cleaning control"—is given appropriate increased display weight.

[0133] Step S240: Generate new advertising content evolution variants containing variant demand features, and supplement them to the original advertising content evolution variant evolution sequence in generational order to form an updated advertising content evolution variant evolution sequence.

[0134] Based on the mutated demand gene fragments and the adjusted generational advertising content evolution variants, a new D3 generation advertising content evolution variant is generated. This variant focuses on the automated cleaning function, applicable scenarios, and user benefits of the smart robotic vacuum cleaner. The D3 generation variant is then added to the original evolutionary sequence in generational order to form an updated advertising content evolution variant evolutionary sequence containing D1, D2, and D3 generations.

[0135] Step S250: Collect user interaction feedback information corresponding to the updated advertising content evolution variant evolution sequence, repeat the above steps, and generate new customized advertising content containing variant demand adaptation features.

[0136] The updated ad content evolution sequence is presented to users, and user interaction feedback information such as browsing time, click history, and feedback is collected. This information is analyzed to determine the user's level of attention and preference for the automated cleaning needs of the smart robot vacuum cleaner. Steps S110 to S150 are repeated to generate new customized ad content that includes features to adapt to the variant needs, such as a description of the smart robot vacuum cleaner's functions, scene adaptation, and experience extension.

[0137] Step S260: Replace the customized advertising content in the original customized advertising content set that conflicts with the variation requirements with the new customized advertising content, and supplement with new customized advertising content that meets the variation requirements to complete the update of the customized advertising content set.

[0138] Examine the existing set of customized ad content. If any content, such as emphasizing "more thorough manual cleaning," conflicts with the changed needs, replace it with new customized ad content. Simultaneously, add newly generated customized ad content containing features adapted to the changed needs to the set, ensuring the customized ad content set fully covers the user's updated needs sequence, thus completing the update.

[0139] Figure 2 The illustration shows exemplary hardware and software components of a user profile-based advertising content customization generation system 100 that can implement the ideas of this application, according to some embodiments of this application. For example, a processor 120 can be used in the user profile-based advertising content customization generation system 100 and to perform the functions in this application.

[0140] For example, the user profile-based advertising content customization generation system 100 may include a network port 110 connected to a network, one or more processors 120 for executing program instructions, a communication bus 130, and various forms of storage media 140, such as a disk, ROM, or RAM, or any combination thereof. Exemplarily, the user profile-based advertising content customization generation system 100 may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The methods of this application can be implemented according to these program instructions. The user profile-based advertising content customization generation system 100 also includes an I / O interface 150 between the computer and other input / output devices.

[0141] Furthermore, this embodiment of the invention also provides a readable storage medium, wherein computer-executable instructions are preset in the readable storage medium, and when the processor executes the computer-executable instructions, the above-mentioned method for customizing and generating advertising content based on user profiles is implemented.

[0142] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A method for customizing and generating advertising content based on user profiles, characterized in that, The method includes: Based on the behavioral preference trajectories and potential demand tendencies in the user profile, the core demand elements with genetic characteristics are decomposed and a demand gene sequence is generated. The demand gene sequence contains multiple evolvable demand segments and the association rules between demand segments. Based on the evolutionary possibilities of each demand segment in the demand gene sequence, an AI algorithm is invoked to generate multiple sets of basic variants of advertising content with differentiated characteristics. Each basic variant of advertising content carries a gene identifier corresponding to the demand segment. The basic variants of the advertising content are subjected to evolutionary iteration processing. Based on the association rules between demand segments in the demand gene sequence, the text structure, visual form and information hierarchy of the basic variants of the advertising content are adjusted to generate multiple generations of advertising content evolution variants, forming an advertising content evolution variant evolution sequence. Collect user interaction feedback information corresponding to the evolutionary sequence of the advertising content evolution variants, identify the advertising content evolution variant with the highest fit with the demand gene sequence based on the user interaction feedback information, and determine the optimal evolutionary path; The advertising content evolution variants are optimized along the optimal evolution path, and the core adaptation features in the advertising content evolution variants are integrated to generate a set of customized advertising content that is deeply compatible with the user profile's needs. The process involves extracting core demand elements with genetic characteristics from user profiles based on behavioral preference trajectories and potential demand tendencies, generating demand gene sequences, including: The behavioral preference trajectory is extracted from the user profile. The behavioral preference trajectory includes the user's content browsing trajectory, product selection trajectory, and service usage trajectory in different periods. Each trajectory carries a time sequence marker and a description of the intensity of the behavior. Extract potential demand tendencies from user profiles. These potential demand tendencies include descriptions of users' extended needs for existing content, descriptions of their needs to explore unexplored areas, and descriptions of their needs to upgrade known services. The behavioral preference trajectory is fragmented and key behavioral segments are selected from high to low according to the behavioral intensity description. Each key behavioral segment corresponds to a potential demand point. The potential demand tendencies are extracted by refining the elements, and the core demand attributes are extracted, including the functional orientation, scenario orientation and experience orientation of the demand. By binding key behavioral fragments with core requirement attributes, core requirement elements are formed, and each core requirement element contains the basis for the behavior and the direction of the requirement. Based on the time sequence of behavioral preference trajectories, the core elements of demand are sorted, and evolutionary association rules between adjacent core elements of demand are defined. The sorted core elements of demand and the evolutionary association rules are integrated to generate a demand gene sequence. The evolutionary association rules are used to describe the process by which the previous core element of demand guides the change of the next core element of demand. Based on the evolutionary probability of each demand segment in the demand gene sequence, an AI algorithm is invoked to generate multiple sets of basic variants of advertising content with differentiated characteristics. Each basic variant of advertising content carries a gene identifier corresponding to the demand segment, including: Analyze each demand segment in the demand gene sequence, extract the core demand attributes and evolutionary association rules of each demand segment, and determine the content direction that each demand segment can be extended to. The core requirement attributes of each requirement fragment are input into the AI ​​algorithm. Based on the functional orientation of the requirement attributes, a draft of the text content is generated; based on the scenario orientation of the requirement attributes, a draft of the visual style is generated; and based on the experience orientation of the requirement attributes, a draft of the information hierarchy is generated. The initial draft of the text content was differentiated by changing the sentence structure, replacing core descriptive words, and adjusting the tone and style to generate multiple sets of advertising content text variations. Each set of advertising content text variations maintained consistency with the core requirement attributes of the requirement fragment. The initial visual style draft was differentiated by changing the color scheme, adjusting the shape of graphic elements, and transforming the layout structure to generate multiple sets of visual variations of the advertising content. Each set of visual variations of the advertising content meets the basic requirements of the scene to which the demand segment is directed. The initial draft of the information hierarchy was differentiated by changing the information display order, adjusting the weight of information modules, and changing the rhythm of information delivery to generate multiple sets of advertising content information variations. Each set of advertising content information variations met the core standards for the experience of the required segment. Each set of text variations, visual variations, and information variations of advertising content is combined to form multiple sets of basic advertising content variations. A gene identifier matching the corresponding demand fragment is added to each basic advertising content variation. The gene identifier includes the demand fragment number and the core demand attribute code.

2. The method for customizing and generating advertising content based on user profiles according to claim 1, characterized in that, The process of iteratively evolving the basic variants of the advertising content involves adjusting the text structure, visual form, and information hierarchy of the basic variants based on the association rules between demand segments in the demand gene sequence, generating multiple generations of advertising content evolution variants, and forming an advertising content evolution variant evolution sequence, including: Extract the evolutionary association rules between adjacent demand segments in the demand gene sequence, and determine the evolutionary information of the transition from the basic variant of advertising content corresponding to the previous demand segment to the basic variant of advertising content corresponding to the next demand segment; Starting with the first-generation basic variant of advertising content, the text structure of the previous basic variant of advertising content is adjusted according to the guiding relationship in the evolutionary association rules. While retaining the core information of the previous basic variant of advertising content text, text description elements corresponding to the next demand fragment are added to form the second-generation advertising content text evolution variant. Adjust the visual form of the previous basic variant of the advertising content, and while maintaining the visual style consistency of the previous basic variant of the advertising content, incorporate the visual element features corresponding to the next demand segment to form the second-generation visual evolution variant of the advertising content. Adjust the information hierarchy of the previous basic variant of the ad content. While maintaining the information transmission logic of the previous basic variant of the ad content, increase the display weight of the information module corresponding to the next demand segment, and adjust the arrangement order of the information modules to guide users to pay attention to the content of the next demand segment, thus forming the second-generation ad content information evolution variant. The second-generation advertising content text evolution variant, the second-generation advertising content visual evolution variant, and the second-generation advertising content information evolution variant are combined to form the second-generation advertising content evolution variant. An evolution generation identifier and corresponding demand fragment association code are added to the second-generation advertising content evolution variant. Repeat the above adjustment steps to generate a third-generation advertising content evolution variant based on the second-generation advertising content evolution variant, and so on, until a multi-generation advertising content evolution variant covering all demand segments of the demand gene sequence is generated, and arranged in generational order to form an advertising content evolution variant evolution sequence.

3. The method for customizing and generating advertising content based on user profiles according to claim 2, characterized in that, The process involves adjusting the text structure of the previous basic variant of the advertising content based on the guiding relationship in the evolutionary association rules. While retaining the core information of the previous basic variant, textual description elements corresponding to the next demand fragment are added to form the second-generation advertising content text evolution variant, including: Extract the core information statements from the basic variant text structure of the previous advertisement content. The core information statements include the functional description, user benefit description and scenario application description of the previous requirement fragment. The guiding relationship in the evolutionary association rules is analyzed to determine the association node between the core information of the previous requirement segment and the content of the next requirement segment. The association node is a functional attribute or scenario attribute that is common to the previous requirement segment and the next requirement segment. Using related nodes as connection points, add transitional statements after the core information statements of the previous advertisement content variation. The transitional statements should clearly define the connection relationship between the related nodes and the content of the next requirement segment. Extract the text description element corresponding to the next requirement fragment. The text description element includes the functional extension description, new scenario application description and added value introduction of the next requirement fragment. Integrate the transition statement with the text description element of the next requirement segment and add it to the text structure of the basic variant of the previous ad content; Redundant information is removed from the basic variant text structure of the adjusted previous advertisement content, and expressions that conflict with or repeat with the content of the subsequent demand segment are deleted. The integrity of the core information of the previous demand segment and the newly added descriptions of the subsequent demand segment is retained, forming the second-generation advertisement content text evolution variant.

4. The method for customizing and generating advertising content based on user profiles according to claim 1, characterized in that, The process of collecting user interaction feedback information corresponding to the evolutionary sequences of the advertising content evolutionary variants, identifying the advertising content evolutionary variants with the highest fit to the demand gene sequences based on the user interaction feedback information, and determining the optimal evolutionary path includes: The advertising content evolution variants in the evolutionary sequence are displayed in generational order. User interaction feedback information for each advertising content evolution variant is collected. The user interaction feedback information includes the user's browsing time record, content click record, and feedback record for the advertising content evolution variant. Each record carries the generational identifier and gene identifier of the corresponding advertising content evolution variant. User interaction feedback information is classified and statistically analyzed. The average browsing time, core content click rate and positive feedback ratio of each generation of advertising content evolution variants are statistically analyzed according to generational identifiers. The interaction data of advertising content evolution variants corresponding to demand segments are statistically analyzed according to gene identifiers. Analyze the relationship between the interaction data and the demand gene sequence of each generation of advertising content evolution variants. When the average browsing time of any generation of advertising content evolution variants is higher than that of other generations of advertising content evolution variants, the core demand attribute of the core content click rate corresponds to the core demand segment, and the positive feedback ratio includes the recognition description of demand extension, mark that generation as a candidate suitable generation. Extract the core features of each advertising content evolution variant in the candidate adaptation generation. The core features include the description of related nodes in the text structure of the advertising content evolution variant, the element fusion method in the visual form of the advertising content evolution variant, and the weight allocation logic in the information hierarchy of the advertising content evolution variant. By comparing the degree of fit between the core features of each advertising content evolution variant in the candidate adaptation generations and the demand gene sequence, the advertising content evolution variant with the best interaction data that completely matches the association rules of the core features and demand fragments is selected as the optimal adaptation advertising content evolution variant. Tracing the generation process of the optimal adapted advertising content evolution variant, determining the evolutionary adjustment steps of the optimal adapted advertising content evolution variant from the basic advertising content variant to the current generation, sorting out the adjustment logic in generational order, forming the optimal evolution path, which includes the adjustment direction, related nodes and core feature retention strategy of each generation of advertising content evolution variant.

5. The method for customizing and generating advertising content based on user profiles according to claim 4, characterized in that, The analysis examines the correlation between interaction data and demand gene sequences of each generation of advertising content evolution variants. When the average viewing time of any generation of advertising content evolution variants is higher than that of other generations, the core demand attribute of the core content click-through rate corresponds to the core demand segment, and the proportion of positive feedback includes acknowledgment of demand extensions, that generation is marked as a candidate suitable generation, including: Calculate the average viewing time of all generational ad content evolution variants, determine the baseline value of the average viewing time, and when the average viewing time of a certain generation of ad content evolution variants exceeds the baseline value by a certain proportion, the generation corresponding to that generation of ad content evolution variants is included in the initial screening of generations. The click-through rate of the core content of the initially screened generational advertising content evolution variants is analyzed. The core demand attributes of the demand fragment corresponding to the core content of each advertising content evolution variant are extracted. The proportion of click behavior concentrated in the description area of ​​the core demand attribute is counted. When the proportion exceeds the preset proportion, it is determined that the click-through rate of the advertising content evolution variant of that generation matches the core demand attribute of the demand fragment. Collect positive feedback on the evolutionary variants of advertising content across generations through preliminary screening. Extract the number of feedback opinions containing descriptions of demand extension recognition. Calculate the proportion of this number to the total feedback opinions. When the proportion exceeds a preset threshold, it is determined that the evolutionary variant of advertising content for that generation has obtained demand extension recognition. Generations that simultaneously meet the average browsing time target, core content click-through rate matching requirement fragment core requirement attributes, and positive feedback including extended requirement recognition descriptions are marked as candidate compatible generations. Candidate adaptation generations are prioritized and ranked from highest to lowest based on a comprehensive score that includes the average viewing time of the ad content evolution variants within each candidate adaptation generation, the click-through rate of the core content, and the proportion of recognition of extended needs. The higher the comprehensive score, the higher the priority of the candidate adaptation generation.

6. The method for customizing and generating advertising content based on user profiles according to claim 1, characterized in that, The process of optimizing ad content evolution variants along the optimal evolution path, integrating core adaptation features from these variants, and generating a set of customized ad content deeply aligned with user profile needs includes: The adjustment steps of each generation of advertising content evolution variants in the optimal evolution path are analyzed, and the core adaptation features retained in each generation of advertising content evolution variants are extracted. The core adaptation features include the related node descriptions at the text level of advertising content evolution variants, the element fusion style at the visual level of advertising content evolution variants, and the weight allocation pattern at the information level of advertising content evolution variants. The extracted core adaptation features are categorized and integrated, and feature sets are established according to three dimensions: text, visual, and information. Each feature set contains the core adaptation features of the advertising content evolution variants of each generation in the optimal evolution path. Analyze the feature relationships in the feature sets of each dimension, and identify the style matching relationship between text features and visual features, the display synergy relationship between visual features and information features, and the logical support relationship between information features and text features; Based on the aforementioned feature correlation, the feature sets of each dimension are optimized and integrated to form text-optimized feature groups, visual-optimized feature groups, and information-optimized feature groups. By applying text optimization feature groups to the creation of advertising content, multiple sets of customized advertising texts that meet the feature requirements are generated. Each set of customized advertising texts includes related node descriptions and extended descriptions of requirements. The visual optimization feature set is applied to the visual design of the advertising content to generate multiple sets of customized advertising visuals that meet the feature requirements. Each set of customized advertising visuals includes elements that blend style and adapt to the required scenario. The information optimization feature group is applied to the information arrangement of the advertising content to generate multiple sets of customized advertising information levels that meet the feature requirements. Each set of customized advertising information levels includes a weight allocation mode and a strategy to highlight the key needs. By combining customized advertising text, customized advertising visuals, and customized advertising information layers, multiple sets of complete customized advertising content are formed. By integrating all the combination schemes, a set of customized advertising content that deeply matches the user profile's needs is generated.

7. The method for customizing and generating advertising content based on user profiles according to claim 6, characterized in that, The method of applying text optimization feature sets to the text creation of advertising content generates multiple sets of customized advertising texts that meet the feature requirements, including: Extract the associated node expression template and the requirement extension description framework from the text optimization feature group. The associated node expression template contains the sentence structure connecting different requirement fragments, and the requirement extension description framework contains the expression logic of function extension, scenario extension, and experience extension. The core demand line of the customized advertising text is determined based on the main demand segments of the demand gene sequence and the evolutionary association rules. Based on the core demand line, select matching related node expression templates, fill the core information of adjacent demand segments into the related node expression templates, and form connecting paragraphs of customized advertising text. Based on the requirement extension description framework, add requirement extension description content after the connecting paragraph; Adjust the overall tone and style of the customized advertising text to match the language preference characteristics in the user profile; Perform a logical coherence check on the content of the customized advertising text to ensure that the related node descriptions, extended descriptions of needs, and the core needs form a complete logical chain, and generate a set of customized advertising texts. By changing the associated node description templates and demand extension description frameworks, repeat the above steps to generate multiple sets of customized advertising texts with differentiated descriptions.

8. A user profile-based advertising content customization and generation system, characterized in that, The device includes a processor and a memory, the memory being connected to the processor. The memory is used to store programs, instructions, or code, and the processor is used to execute the programs, instructions, or code in the memory to implement the user profile-based advertising content customization generation method according to any one of claims 1-7.

Images