Ad creative design method and system based on ad sample analysis

Abstract

The application relates to the field of advertisement creative design, and discloses an advertisement creative design method and system based on advertisement sample analysis. The method comprises the following steps: obtaining multiple competitor video advertisements of a target product, and obtaining key frame screenshots, positive comment times and positive comment keyword word clouds of high complete play rate segments corresponding to the multiple competitor video advertisements respectively; clustering the multiple competitor video advertisements according to the key frame screenshots, the positive comment times and the positive comment keyword word clouds of the high complete play rate segments corresponding to the multiple competitor video advertisements respectively, obtaining multiple competitor video advertisement categories, and determining a target competitor video advertisement category in the multiple competitor video advertisement categories; and determining an advertisement creative design scheme corresponding to the target product according to the multiple competitor video advertisements in the target competitor video advertisement category through an advertisement creative design model. The application can automatically analyze and design with pertinence the strategies, advantages and innovative points of competitors in advertisement creativity.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of advertisement creative design, and more particularly to an advertisement creative design method and system based on advertisement sample analysis. BACKGROUND

[0002] In the development process of the advertising industry, the traditional advertisement creative design method mainly depends on the personal experience, inspiration and basic market research of the designer. For example, the advertisement designer may generate creative schemes through brainstorming and other ways according to the experience of past successful cases, combined with the general understanding of the target audience group, and then screen and optimize the schemes. This design mode based on experience and inspiration has provided strong support for the development of the advertising industry in a certain historical period, and has helped many enterprises to complete the advertising promotion task.

[0003] However, the traditional advertisement creative design method has a significant defect, that is, it can only design advertisements according to the experience of the designer and the advertisement creativity of the competitors. In the traditional mode, the designer often lacks the real-time acquisition, in-depth analysis and effective utilization mechanism of the advertisement creativity of the competitors. They focus more on the characteristics of their own brand and the target audience, and can only analyze and design according to the past experience and the strategies, advantages and innovative points of the competitors in the advertisement creativity, which leads to the homogenization of the advertisement creativity of the enterprise and the competitors, and the inability to accurately highlight the differentiated advantages, so as to be difficult to stand out in the fierce market competition and to fully meet the higher demand of the enterprise for the advertisement creative design in the competitive environment and the information development. SUMMARY

[0004] The purpose of the present application is to provide an advertisement creative design method and system based on advertisement sample analysis, which solves the technical problem that the strategies, advantages and innovative points of the competitors in the advertisement creativity can only be analyzed and designed according to the past experience, and achieves the technical effect that the strategies, advantages and innovative points of the competitors in the advertisement creativity are automatically analyzed and designed.

[0005] The embodiment of the application provides an advertisement creative design method based on advertisement sample analysis, which comprises the following steps: obtaining a plurality of competitive product video advertisements of a target product, and obtaining key frame screenshots, positive comment times and positive comment keyword word clouds of high complete play rate segments corresponding to the plurality of competitive product video advertisements respectively; clustering the plurality of competitive product video advertisements according to the key frame screenshots, the positive comment times and the positive comment keyword word clouds of the high complete play rate segments corresponding to the plurality of competitive product video advertisements respectively, obtaining a plurality of competitive product video advertisement categories, and determining a target competitive product video advertisement category in the plurality of competitive product video advertisement categories; and determining an advertisement creative design scheme corresponding to the target product according to the plurality of competitive product video advertisements in the target competitive product video advertisement category through an advertisement creative design model.

[0006] In a possible implementation, the clustering of the plurality of competitive product video advertisements according to the key frame screenshots, the positive comment times and the positive comment keyword word clouds of the high complete play rate segments corresponding to the plurality of competitive product video advertisements respectively, and the obtaining of the plurality of competitive product video advertisement categories, comprise the following steps: determining advertisement segment features corresponding to the plurality of competitive product video advertisements respectively according to the key frame screenshots of the high complete play rate segments corresponding to the plurality of competitive product video advertisements respectively through a key frame identification model corresponding to the target product, wherein the advertisement segment features represent image content features of the key frame screenshots of the high complete play rate segments; determining a causal relationship feature according to the time stamps corresponding to the key frame screenshots, the positive comment times and the positive comment keyword word clouds of the high complete play rate segments through a causal relationship derivation model, wherein the causal relationship feature represents a causal relationship between the key frame screenshots of the high complete play rate segments and the positive comment times and the positive comment keyword word clouds; and clustering the plurality of competitive product video advertisements according to the advertisement segment features and the causal relationship feature corresponding to the plurality of competitive product video advertisements respectively, and obtaining the plurality of competitive product video advertisement categories.

[0007] In another possible implementation, the clustering of the plurality of competitive product video advertisements according to the key frame screenshots, the positive comment times and the positive comment keyword word clouds of the high complete play rate segments corresponding to the plurality of competitive product video advertisements respectively, and the obtaining of the plurality of competitive product video advertisement categories, further comprise the following steps: determining product similarities between the competitive products corresponding to the plurality of competitive product video advertisements respectively and the target product; and clustering the plurality of competitive product video advertisements according to the product similarities, the advertisement segment features and the causal relationship feature corresponding to the plurality of competitive product video advertisements respectively, and obtaining the plurality of competitive product video advertisement categories.

[0008] In another possible implementation, the plurality of competitive video advertisements are clustered according to the key frame screenshots, positive comment times, and positive comment keyword word clouds of the high complete play rate segments respectively corresponding to the plurality of competitive video advertisements, to obtain a plurality of competitive video advertisement categories, and further comprising: determining core selling point similarity, packaging method similarity, and delivery strategy similarity of the target product and the competitive product respectively corresponding to the plurality of competitive video advertisements, and determining the sum of the inverses of the core selling point similarity, the inverse of the packaging method similarity, and the inverse of the delivery strategy similarity of the competitive product respectively corresponding to the plurality of competitive video advertisements as competitive factor difference degrees respectively corresponding to the plurality of competitive video advertisements; and clustering the plurality of competitive video advertisements according to the competitive factor difference degrees respectively corresponding to the plurality of competitive video advertisements, the advertisement segment features, and the causal relationship features, to obtain the plurality of competitive video advertisement categories.

[0009] In another possible implementation, the target competitive video advertisement category in the plurality of competitive video advertisement categories is determined, comprising: obtaining user comment density, positive comment sentiment index, and complete play growth rate index of the competitive video advertisements in the plurality of competitive video advertisement categories within a preset time period; determining advertisement performance evaluation values respectively corresponding to the plurality of competitive video advertisement categories according to the user comment density, the positive comment sentiment index, and the complete play growth rate index of the competitive video advertisements in the plurality of competitive video advertisement categories within the preset time period; and determining a maximum advertisement performance evaluation value in the advertisement performance evaluation values respectively corresponding to the plurality of competitive video advertisement categories, and taking the competitive video advertisement category corresponding to the maximum advertisement performance evaluation value as the target competitive video advertisement category.

[0010] In another possible implementation, the target competitive video advertisement category in the plurality of competitive video advertisement categories is determined, further comprising: obtaining a target budget factor corresponding to the target product, and obtaining a budget factor corresponding to the competitive video advertisements in the plurality of competitive video advertisement categories; determining a difference value between the budget factor corresponding to the competitive video advertisements in the plurality of competitive video advertisement categories and the target budget factor as a budget factor difference value; wherein the budget factor represents the budget amount of the video advertisement; determining the sum of the advertisement performance evaluation values respectively corresponding to the plurality of competitive video advertisement categories and the budget factor difference value as an advertisement category evaluation value; and determining a maximum advertisement category evaluation value in the advertisement category evaluation values respectively corresponding to the plurality of competitive video advertisement categories, and taking the competitive video advertisement category corresponding to the maximum advertisement category evaluation value as the target competitive video advertisement category.

[0011] In another possible implementation, the determining the target competitor video advertisement category from the plurality of competitor video advertisement categories further includes: obtaining a target brand style feature value corresponding to the target product, and obtaining brand style feature values corresponding to the competitor video advertisements in the plurality of competitor video advertisement categories; determining a difference between the brand style feature values corresponding to the competitor video advertisements in the plurality of competitor video advertisement categories and the target brand style feature value as a brand style difference value; wherein the brand style feature value represents a low-price promotion feature and a luxury feature of the video advertisement; determining a sum of the advertisement performance evaluation value, the budget factor difference value, and the brand style difference value corresponding to the competitor video advertisement categories respectively as an advertisement category evaluation value; and determining a maximum advertisement category evaluation value in the advertisement category evaluation values corresponding to the plurality of competitor video advertisement categories as the target competitor video advertisement category.

[0012] In another possible implementation, the determining the target competitor video advertisement category from the plurality of competitor video advertisement categories further includes: obtaining a historical style evaluation value of a historical advertisement of the target product, and obtaining style evaluation values of the competitor video advertisements in the plurality of competitor video advertisement categories; determining a difference between the style evaluation values of the competitor video advertisements in the plurality of competitor video advertisement categories and the historical style evaluation value as an advertisement style innovation value; wherein the style evaluation value includes style evaluation feature values in multiple dimensions; determining a sum of the advertisement performance evaluation value, the budget factor difference value, the brand style difference value, and the advertisement style innovation value corresponding to the competitor video advertisement categories respectively as an advertisement category evaluation value; and determining a maximum advertisement category evaluation value in the advertisement category evaluation values corresponding to the plurality of competitor video advertisement categories as the target competitor video advertisement category.

[0013] In another possible implementation, the determining the target competitor video advertisement category from the plurality of competitor video advertisement categories further includes: obtaining a negative evaluation rate corresponding to the competitor video advertisements in the plurality of competitor video advertisement categories; determining a quotient of the advertisement category evaluation value and the negative evaluation rate corresponding to the competitor video advertisement categories respectively and normalizing the quotient as an advertisement category evaluation adjustment value; and determining a maximum advertisement category evaluation adjustment value in the advertisement category evaluation adjustment values corresponding to the plurality of competitor video advertisement categories as the target competitor video advertisement category.

[0014] The embodiment of the present application further provides an advertisement creative design system based on advertisement sample analysis, which includes units for executing the method according to any one of the above.

[0015] Compared with the prior art, the embodiment of the present application has the beneficial effects that:

[0016] The embodiment of the present application provides an advertisement creative design method based on advertisement sample analysis, and the method comprises the following steps: obtaining a plurality of competitive product video advertisements of a target product, and obtaining key frame screenshots, positive comment times and positive comment keyword word clouds of high complete play rate segments corresponding to the plurality of competitive product video advertisements respectively; clustering the plurality of competitive product video advertisements according to the key frame screenshots, the positive comment times and the positive comment keyword word clouds of the high complete play rate segments corresponding to the plurality of competitive product video advertisements respectively, obtaining a plurality of competitive product video advertisement categories, and determining a target competitive product video advertisement category in the plurality of competitive product video advertisement categories; and determining an advertisement creative design scheme corresponding to the target product according to the plurality of competitive product video advertisements in the target competitive product video advertisement category through an advertisement creative design model. The embodiment of the present application can determine the advertisement creative design scheme corresponding to the target product according to the plurality of competitive product video advertisements in the target competitive product video advertisement category, and can enhance the attraction of the advertisement to target users. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0018] Figure 1 The flowchart of the first advertisement creative design method based on advertisement sample analysis provided by the embodiment of the present application is shown in the figure.

[0019] Figure 2 The working flowchart of the first advertisement creative design method based on advertisement sample analysis provided by the embodiment of the present application is shown in the figure.

[0020] Figure 3 The flowchart of the second advertisement creative design method based on advertisement sample analysis provided by the embodiment of the present application is shown in the figure.

[0021] Figure 4 The flowchart of the third advertisement creative design method based on advertisement sample analysis provided by the embodiment of the present application is shown in the figure.

[0022] Figure 5 The logic structure diagram of the first advertisement creative design system based on advertisement sample analysis provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0023] It should be understood that the word “comprise” or variations such as “comprises” or “comprising”, when used in this specification and in the accompanying claims, specify the presence of stated features, integers, steps, operations, elements, components and / or groups of features, integers, steps, operations, elements, components but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0024] It should also be understood that the term “and / or” when used in this specification and in the following claims is to be interpreted as an inclusive-inclusive rather than an exclusive-inclusive. That is, unless it is otherwise specified, “and / or” should be interpreted to mean one and only one of the items in the list is present or one or more of the items in the list are present.

[0025] As used in this specification and claims, the terms “if’ and “when” can be interpreted to mean “upon determination” or “in response to a determination” or “upon detection” or “in response to a detection” depending on the context. Similarly, the phrase “if determined” or “if detected [the described condition or event]” can be interpreted to mean “upon determination” or “in response to a determination” or “upon detection” or “in response to a detection” of [the described condition or event] depending on the context.

[0026] In addition, the terms “first”, “second”, “third”, etc. as used in the description of the specification and the appended claims are only used to distinguish descriptions and cannot be understood as indicating or implying relative importance.

[0027] Reference in the specification to “one embodiment” or “some embodiments” etc. means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. The appearances of the phrases “in one embodiment”, “in some embodiments”, “in other embodiments”, “in additional embodiments” etc. in various places in the specification are not necessarily all referring to the same embodiment, although they can. The terms “comprise”, “comprises”, “comprising”, “include”, “includes”, “including” and the like are synonymous with “containing” or “comprising” and are used in the sense of “including but not limited to”, unless otherwise stated.

[0028] In the traditional creative design method of advertising, the strategy, advantage and innovation point of competitors in the creative design of advertising can only be analyzed and designed according to the past experience.

[0029] Based on the above reasons, this application provides an advertising creative design method based on advertising example analysis. The method includes: acquiring multiple competitor video ads for a target product, and acquiring keyframe screenshots, positive comment times, and positive comment keyword clouds of high completion rate segments corresponding to each of the multiple competitor video ads; clustering the multiple competitor video ads based on the keyframe screenshots, positive comment times, and positive comment keyword clouds of the high completion rate segments corresponding to each of the multiple competitor video ads to obtain multiple competitor video ad categories, and determining a target competitor video ad category among the multiple competitor video ad categories; and determining an advertising creative design scheme corresponding to the target product based on the multiple competitor video ads in the target competitor video ad category using an advertising creative design model. This application can determine a suitable advertising creative design scheme for the target product based on multiple competitor video ads in the target competitor video ad category, thereby enhancing the attractiveness of the advertisement to the target user.

[0030] In some scenarios, the advertising creative design method based on advertising sample analysis of this application embodiment can be applied to the advertising creative design methods of various daily consumer goods, electronic products and luxury goods, which can improve the speed and efficiency of advertising creative design methods.

[0031] The following specific examples illustrate an advertising creative design method based on advertising sample analysis provided in this application.

[0032] Figure 1 A flowchart illustrating the first advertising creative design method based on advertising sample analysis provided in this application embodiment is shown below. Figure 1 As shown, this advertising creative design method based on advertising sample analysis includes S110 to S130, and S110 to S130 will be explained in detail below.

[0033] S110. Obtain multiple competitor video ads for the target product, and obtain keyframe screenshots, positive comment times, and positive comment keyword clouds for the high completion rate segments corresponding to each competitor video ad.

[0034] Figure 2 A schematic diagram of the workflow of the first advertising creative design method based on advertising sample analysis provided in the embodiments of this application is shown below. Figure 2 As shown, in the advertising creative design method based on advertising sample analysis in this implementation, multiple competitor video ads for the target product can be obtained first, and keyframe screenshots, positive comment times, and positive comment keyword clouds of high completion rate segments can be extracted from these competitor video ads.

[0035] When designing advertising creatives, keyframe screenshots can capture the most frequently viewed and exciting scenes in the advertisement, revealing the visual design that attracts user interest.

[0036] In the process of designing the advertisement creative, the positive comment time is used to identify the location of the user's positive feedback, which can specifically identify the advertisement climax part in the video advertisement.

[0037] In the process of designing the advertisement creative, the positive comment keyword word cloud can aggregate common positive words in the comments, such as "creative" or "fun", highlighting the highlights of the advertisement.

[0038] After obtaining the key frame screenshot, the positive comment time and the positive comment keyword word cloud, the key frame screenshot, the positive comment time and the positive comment keyword word cloud can provide comprehensive advertisement effect indicators for subsequent steps, helping to identify the core features of the competitor video advertisements.

[0039] S120, according to the key frame screenshot, the positive comment time and the positive comment keyword word cloud of the high complete playback rate segment corresponding to each of the plurality of competitor video advertisements, clustering the plurality of competitor video advertisements to obtain a plurality of competitor video advertisement categories, and determining a target competitor video advertisement category in the plurality of competitor video advertisement categories.

[0040] As shown in Figure 2 After obtaining the key frame screenshot, the positive comment time and the positive comment keyword word cloud of the high complete playback rate segment, the plurality of competitor video advertisements can be clustered and grouped according to the data characteristics of these competitor video advertisements, including the key frame screenshot, the positive comment time and the positive comment keyword word cloud of the high complete playback rate segment, to obtain a plurality of competitor video advertisement categories.

[0041] After obtaining the plurality of competitor video advertisement categories, the plurality of competitor video advertisement categories can reveal similar theme styles or user preference patterns between advertisements, thereby determining the most relevant target competitor video advertisement category.

[0042] For example, in the analysis process, unsupervised learning algorithms can be used to group competitor video advertisements, where some categories may focus on emotional marketing, while others emphasize product functions, which can quickly filter out target categories that are highly matched with the target product, and optimize the design direction of the video advertisement.

[0043] S130, determining the advertisement creative design scheme corresponding to the target product according to the plurality of competitor video advertisements in the target competitor video advertisement category through the advertisement creative design model.

[0044] As shown in Figure 2As shown, after obtaining multiple competitor video advertisement categories, an advertisement creative design model can be used to generate an advertisement creative design scheme corresponding to the target product based on multiple competitor video advertisements in the target competitor video advertisement category. When designing the advertisement creative, the advertisement creative design model can integrate the clustered category features, such as key frame visual elements and positive review keyword trends, to innovatively design a new advertisement concept.

[0045] For example, the advertisement creative design model can learn the high completion rate screenshot patterns of the target category advertisements, and combine the popular themes of the keyword clouds, such as “clever plot” or “visual impact”, to plan the narrative structure of the video advertisement, and ensure that the design scheme is based on real user feedback and competitor advantages.

[0046] For example, the advertisement creative design model can be a Large Language Model (LLM) model fine-tuned based on multiple sample competitor video advertisements in the sample competitor video advertisement category and a sample advertisement creative design scheme.

[0047] The above implementation has the beneficial effects of efficiently identifying advertisement trends through clustering analysis, reducing blind trial and error in the design process, and improving the accuracy and speed of creative generation; and determining an advertisement creative design scheme suitable for the target product based on multiple competitor video advertisements in the target competitor video advertisement category through the advertisement creative design model, thereby enhancing the appeal of the advertisement to target users.

[0048] The above implementation also has the beneficial effects of improving the accuracy of clustering analysis of competitor advertisements by combining visual content and user feedback keyword clouds, and taking into account the visual effect and communication effect of the advertisement.

[0049] In some implementations, in S120, the multiple competitor video advertisements are clustered based on the key frame screenshots of the high completion rate segments, the positive review times, and the positive review keyword clouds of the multiple competitor video advertisements respectively, to obtain multiple competitor video advertisement categories, including S121-S122, which are described below.

[0050] S121, through a target product corresponding key frame identification model, determines advertisement segment features corresponding to the multiple competitor video advertisements respectively based on the key frame screenshots of the high completion rate segments of the multiple competitor video advertisements respectively, the advertisement segment features representing image content features of the key frame screenshots of the high completion rate segments. Through a causal relationship derivation model, a causal relationship feature is determined based on the time stamps corresponding to the key frame screenshots of the high completion rate segments, the positive review times, and the positive review keyword clouds, the causal relationship feature representing the causal relationship between the key frame screenshots of the high completion rate segments and the positive review times and the positive review keyword clouds.

[0051] In the advertisement creative design process based on the advertisement example analysis, the key frame screenshots of the high complete play rate segments of the plurality of competitive video advertisements can be processed by the key frame recognition model special for the target product. Through the key frame recognition model, the visual content features contained in the key frame screenshots can be extracted, and the advertisement segment features representing the core picture elements of the advertisements are generated.

[0052] For example, when analyzing the video advertisements of the sports shoes category, the key frame recognition model corresponding to the target product of the sports shoes category can identify the image features of the slow-motion display of the shoe sole elasticity in the high complete play rate segments, and capture the core visual design elements that attract the attention of users.

[0053] Exemplarily, the key frame recognition model can be a multi-modal based visual understanding model, and the key frame recognition model is used to extract the visual content features contained in the key frame screenshots.

[0054] After obtaining the advertisement segment features, the causal relationship derivation model can be used to analyze the timestamp positions corresponding to the key frame screenshots, the occurrence times of the positive comments, and the keyword word clouds of the positive comments. The causal relationship derivation model can establish the association rules between the key frame picture content and the positive feedback of the users, and generate the feature vectors representing the causal relationship among the three.

[0055] For example, when the product close-up key frame appears at 15 seconds in the video advertisement, the causal relationship derivation model can deduce that there is a strong association between the product close-up key frame and the positive comment keywords such as “burning” appearing at the 20 second time point, which indicates that the close-up picture can trigger the expression of the audience's excited emotion, and then the causal relationship between the advertisement effects contained in the video advertisement is mined through the feature vectors representing the causal relationship among the three.

[0056] Exemplarily, the causal relationship derivation model can be a neural network based deep learning model, and the causal relationship derivation model can be trained by the sample timestamp positions corresponding to the sample key frame screenshots, the sample positive comment occurrence times, the sample positive comment keyword word clouds, and the sample causal relationship features.

[0057] S122, according to the advertisement segment features and the causal relationship features corresponding to the plurality of competitive video advertisements respectively, the plurality of competitive video advertisements are clustered to obtain a plurality of competitive video advertisement categories.

[0058] After obtaining the advertisement segment features and the causal relationship features corresponding to the plurality of competitive video advertisements respectively, the advertisement segment features and the causal relationship features can be fused, and the joint clustering analysis of the plurality of competitive video advertisements is performed, and then the advertisements can be divided into categories with similar successful elements according to the composite association mode of the visual content and the user feedback.

[0059] For example, based on visual features, the sports shoe advertisements are divided into a "technology display" group and a "scene experience" group, and meanwhile, the causal relationship features are combined to find that the "technology display" group is associated with comments such as "innovation" and "light", and the "scene experience" group is associated with keywords such as "comfort" and "fashion". Through further clustering analysis, a multi-dimensional classification of video advertisements is formed.

[0060] The implementation manner has the beneficial effects that the dual feature extraction mechanism can deeply analyze the internal relationship of the success factors of the advertisements, and the clustering results can reflect the coupling relationship between the content design and the user feedback.

[0061] The implementation manner also has the beneficial effects that the causal relationship features are used as the classification basis, which can accurately distinguish the types of advertisements that are similar in appearance but different in user response mechanism, provide a high-value reference dimension for subsequent creative design, and improve the effect of the creative design of the advertisements.

[0062] In some implementation manners, in the S120, the multiple competitor video advertisements are clustered according to the key frame screenshots, the positive comment times and the positive comment keyword word clouds of the high complete play rate segments corresponding to the multiple competitor video advertisements, and multiple competitor video advertisement categories are obtained. The S123 to S124 are described below.

[0063] The S123 determines the product similarity of the competitors and the target product corresponding to the multiple competitor video advertisements, respectively.

[0064] In the process of the advertisement creative design based on the advertisement sample analysis, the product similarity between the competitors and the target product corresponding to the multiple competitor video advertisements can be determined. The product similarity represents the similarity of the competitors and the target product in the aspects of the category, the function and the structure, and then the product similarity helps to evaluate whether the competitor advertisements have reference value.

[0065] For example, in the analysis of the video advertisements, for a newly released sports shoe target product, the product similarity of the competitor sports shoes can be compared. Specifically, the commonality of the brand positioning and the use scene can be compared, so that the most relevant advertisement sample is selected for subsequent analysis. This ensures that the clustering process is based on the comparability of the products, and avoids including irrelevant advertisements in the grouping.

[0066] The S124 clusters the multiple competitor video advertisements according to the product similarity, the advertisement segment features and the causal relationship features corresponding to the multiple competitor video advertisements, respectively, and obtains multiple competitor video advertisement categories.

[0067] In the implementation mode, the product similarity, the advertisement segment feature and the causality feature can be combined in the cluster analysis to cluster and group the multiple competitor video advertisements, and the interaction of the product similarity and the advertisement feedback data can be considered at the same time, and the accuracy of the category division can be comprehensively optimized.

[0068] For example, the multiple competitor video advertisements can be clustered and grouped according to the product similarity, the advertisement segment feature and the causality feature of a certain competitor sports shoes, and the "professional sports" group and the "leisure fashion" group can be distinguished, and the advertisement success elements and the product correlation mode can be accurately captured.

[0069] The implementation mode has the beneficial effects that the product similarity is introduced as a key variable, irrelevant competitor interference can be filtered, the pertinence and practicality of the clustering result can be improved, and the advertisement creative design can be ensured to be more in line with the target product characteristics.

[0070] The implementation mode also has the beneficial effects that the multi-dimensional data of the product similarity, the advertisement segment feature and the causality feature can enhance the accuracy of the category division, avoid deviation caused by relying on the advertisement performance alone, and improve the efficiency of the creative generation process and the user preference matching degree.

[0071] In some implementation modes, in the S120, the multiple competitor video advertisements are clustered according to the key frame screenshots, the positive comment time and the positive comment keyword word cloud of the high complete play rate segments corresponding to the multiple competitor video advertisements respectively, and the multiple competitor video advertisement categories are obtained, and the S125 to S126 are further included, and the S125 to S126 are specifically described as follows.

[0072] In the S125, the core selling point similarity, the packaging method similarity and the delivery strategy similarity of the competitor and the target product corresponding to the multiple competitor video advertisements respectively are determined, and the sum of the reciprocal of the core selling point similarity, the reciprocal of the packaging method similarity and the reciprocal of the delivery strategy similarity of the competitor corresponding to the multiple competitor video advertisements respectively is determined as the competition factor difference degree of the multiple competitor video advertisements.

[0073] In the implementation mode, the core selling point similarity, the packaging method similarity and the delivery strategy similarity between the competitor and the target product corresponding to the multiple competitor video advertisements can be further determined, and these similarity indexes respectively represent the matching degree of the competitor and the target product in the product function core appeal, the appearance packaging form and the marketing strategy.

[0074] For example, when analyzing pet food video advertisements, the core selling point similarity can measure whether the competitor advertisement also emphasizes the core value of "natural food materials", the packaging method similarity can evaluate whether the packaging visual style is similar, and the delivery strategy similarity can compare whether they all choose to deliver advertisements in pet blogger channels.

[0075] In the implementation mode, by calculating the sum of the reciprocals of the core selling point similarity, the packaging method similarity and the delivery strategy similarity, a competition factor difference degree index representing the degree of competition difference can be generated, and the competition factor difference degrees corresponding to the plurality of competitive product video advertisements respectively, the greater the competition factor difference degree, the greater the difference between the competitive strategy of the competitive product and the target product.

[0076] For example, when a competitive product uses high-end gift box packaging and the target product uses environmentally friendly simple packaging, the decrease in packaging method similarity will lead to an increase in its reciprocal, thereby increasing the competition factor difference degree score of the competitive product, and thereby quantitatively identifying the advertisement cases that have significant competition differences with the target product.

[0077] S126, according to the competition factor difference degrees corresponding to the plurality of competitive product video advertisements respectively, the advertisement segment features and the causal relationship features, clustering the plurality of competitive product video advertisements to obtain a plurality of competitive product video advertisement categories.

[0078] After obtaining the competition factor difference degree, in the clustering process, the competition factor difference degree, the advertisement segment features and the causal relationship features can be further integrated for comprehensive analysis, which can ensure that the market differentiation competition dimensions are fully considered when grouping the advertisements, and avoid grouping advertisements with significant competition strategy differences into the same category.

[0079] For example, when clustering pet food advertisements, vegetarian food advertisements with high competition factor difference degrees will be separated from the mainstream meat food advertisements, and combined with the key frame features of high completion rate segments showing vegetable ingredients and the causal relationship features of the "healthy choice" keywords in user comments, an advertisement category reflecting the competition strategy of the submarket is finally formed.

[0080] The implementation mode has the beneficial effects that the competition factor difference degree index accurately identifies the differentiated competition strategy, the clustering results effectively distinguish between advertisements of similar surface but different market competition positioning, and the risk avoidance ability of creative design is improved.

[0081] The implementation mode also has the beneficial effects that the three-dimensional similarity index can present the market competition situation in three dimensions, avoid strategy misjudgment caused by relying solely on advertisement content features, and enhance the adaptability of creative design schemes to the market competition environment.

[0082] Figure 3 The flowchart of the second advertisement creative design method based on advertisement example analysis provided by the embodiments of the present application is shown in FIG. 2, and the S120 in the above is shown in FIG. 2. Figure 3 S210 to S220, which are specifically described as follows.

[0083] S210, obtain the user comment density, positive comment sentiment index and complete playback growth rate index of the competitor video advertisements in the preset time period in the plurality of competitor video advertisement categories. According to the user comment density, positive comment sentiment index and complete playback growth rate index of the competitor video advertisements in the plurality of competitor video advertisement categories in the preset time period, the advertisement performance evaluation values corresponding to the plurality of competitor video advertisement categories respectively are determined.

[0084] In the implementation manner, the user comment density, positive comment sentiment index and complete playback growth rate index of the competitor video advertisements contained in the plurality of competitor video advertisement categories in the preset time period can be obtained respectively. These indexes can present the comprehensive performance effect of the video advertisements of each advertisement category in the short term in a three-dimensional manner, and then the target competitor video advertisement category most suitable for advertisement creative generation is determined through the comprehensive performance effect in the short term.

[0085] In the implementation manner, the user comment density reflects the discussion heat of the advertisement, such as the number of comments generated per unit time within 72 hours after the video advertisement is published.

[0086] In the implementation manner, the positive comment sentiment index can quantify the proportion of positive emotions of users through a sentiment analysis model.

[0087] In the implementation manner, the complete playback growth rate is used to count the improvement amplitude of the advertisement playback completion rate.

[0088] Exemplarily, the preset time period can be 30 days.

[0089] After the user comment density, positive comment sentiment index and complete playback growth rate index are obtained, the advertisement performance evaluation values corresponding to the plurality of competitor video advertisement categories respectively can be calculated according to the user comment density, positive comment sentiment index and complete playback growth rate index of the plurality of competitor video advertisement categories, and the advertisement performance evaluation values can be generated through weighted fusion of the three-dimensional indexes.

[0090] It should be noted that when the advertisement performance evaluation values are generated through weighted fusion of the three-dimensional indexes, the complete playback growth rate can be given a higher weight, directly reflecting the core attraction of the advertisement, so as to ensure that the evaluation values accurately capture the real effectiveness of the advertisement category.

[0091] S220, determine the maximum advertisement performance evaluation value in the advertisement performance evaluation values corresponding to the plurality of competitor video advertisement categories respectively, and take the competitor video advertisement category corresponding to the maximum advertisement performance evaluation value as the target competitor video advertisement category.

[0092] After obtaining the advertisement performance evaluation values corresponding to the multiple competitor video advertisement categories respectively, the maximum value in the advertisement performance evaluation values of the multiple competitor video advertisement categories can be identified, and the competitor video advertisement category corresponding to the maximum value is determined as the target competitor video advertisement category, so that the advertisement strategy mode with the best performance in the current market environment can be automatically screened.

[0093] For example, when the complete play growth rate of a certain food advertisement category within a preset time period of 30 days is 35%, and the positive comment sentiment index is 0.85, the performance evaluation value of the food advertisement category is significantly higher than that of other categories, and then the combination of “interesting plot + product close-up” included in the food advertisement category can be selected as the reference benchmark for creative design.

[0094] The above-mentioned implementation manner has the beneficial effects that the advertisement strategy effectiveness is quantitatively evaluated by using dynamic indexes, subjective judgment deviation is avoided, and it is ensured that the selected advertisement category has an empirical success basis; the advertisement performance evaluation value reflects the real viewing behavior of users and can avoid data noise interference, and the commercial value conversion potential of the advertisement creative scheme design is improved.

[0095] In some implementation manners, in the S120, the target competitor video advertisement category in the multiple competitor video advertisement categories is determined, including S230 to S240, which are specifically described as follows.

[0096] In S230, a target budget factor corresponding to the target product is obtained, and a budget factor corresponding to a competitor video advertisement in the multiple competitor video advertisement categories is obtained. A difference value between the budget factor corresponding to the competitor video advertisement in the multiple competitor video advertisement categories and the target budget factor is determined as a budget factor difference value. The budget factor represents the budget amount of the video advertisement.

[0097] In the implementation manner, the target budget factor corresponding to the target product can be obtained, and the target budget factor represents the budget input intensity of the target product video advertisement. Meanwhile, the budget factors of the competitor advertisements in the multiple competitor video advertisement categories can be obtained, and the budget factor represents the budget amount (i.e. the budget input intensity) of the video advertisement.

[0098] For example, in the analysis of a sports shoe video advertisement, the target budget factor can be set as a medium input level, and if a certain competitor advertisement category uses a star endorsement and large-scale special effect production, the budget factor of the competitor advertisement category can be significantly higher than the target budget factor. If another competitor category uses a user-generated content mode, the budget factor can be close to or lower than the target budget factor.

[0099] In the implementation manner, the difference value between the budget factor corresponding to the competitor advertisement category and the target budget factor can be calculated as the budget factor difference value, and the budget factor difference value quantifies the budget adaptation degree.

[0100] For example, when a certain smart phone advertisement category adopts a film-level production, the positive difference value reflects the risk of over-investment of resources; when a certain home product advertisement category adopts a group performance live shooting strategy, the negative difference value shows its cost efficiency advantage, thereby enabling the budget factor difference value to objectively evaluate the referenceability of each category scheme in the budget dimension.

[0101] S240, determine the sum of the advertisement performance evaluation value and the budget factor difference value corresponding to each of the plurality of competitive video advertisement categories as the advertisement category evaluation value. Determine the maximum advertisement category evaluation value in the advertisement category evaluation values corresponding to the plurality of competitive video advertisement categories, and take the competitive video advertisement category corresponding to the maximum advertisement category evaluation value as the target competitive video advertisement category.

[0102] After obtaining the budget factor difference value, the sum of the advertisement performance evaluation value and the budget factor difference value can be calculated as the advertisement category evaluation value. The advertisement performance evaluation value is related to the historical effect indicators (such as user comment density and complete playback growth rate), and the budget factor difference value can be further used to correct the performance evaluation value by the pre-budget feature.

[0103] For example, the advertisement performance evaluation value of a certain cosmetic video advertisement category is 0.92, but the budget factor difference value is -0.5, so the advertisement category evaluation value can be determined as 0.42.

[0104] In the present implementation, the maximum value of the advertisement category evaluation value can be finally identified, and the competitive video advertisement category corresponding to the maximum value of the advertisement category evaluation value is taken as the target competitive video advertisement category. This selection mechanism of the target competitive video advertisement category considers both the advertisement effect data and ensures the budget feasibility.

[0105] The implementation mode has the beneficial effects that the advertisement performance evaluation value is dynamically corrected by the budget factor difference value, which can avoid the risk of resource overbudget caused by simply pursuing the advertisement effect, and improve the business feasibility of the creative scheme; the budget adaptation degree and the propagation effect are included in the unified evaluation system, so that the finally selected advertisement category realizes the best balance between the effect maximization and the cost controllability, and the investment return rate expectation of the advertisement placement is enhanced.

[0106] In some implementations, in the S120, determining the target competitive video advertisement category in the plurality of competitive video advertisement categories further includes S250 to S260, which are specifically described as follows.

[0107] S250, obtain the target brand style feature value corresponding to the target product, and obtain the brand style feature value corresponding to the competitor video advertisement in the plurality of competitor video advertisement categories. Determine the difference between the brand style feature value corresponding to the competitor video advertisement in the plurality of competitor video advertisement categories and the target brand style feature value as the brand style difference value. The brand style feature value represents the low-price promotion feature and the luxury feature of the video advertisement.

[0108] In the implementation manner, the target brand style feature value corresponding to the target product can be obtained, and the target brand style feature value represents the performance intensity of the advertisement strategy in the two dimensions of the low-price promotion feature and the luxury feature.

[0109] Meanwhile, the brand style feature value of the competitor advertisement in the plurality of competitor video advertisement categories can be obtained, and the brand style feature value can be obtained by video analysis on the competitor advertisement in the plurality of competitor video advertisement categories.

[0110] For example, for a car video advertisement, if the target brand is positioned as an economic family car, the target brand style feature value can be biased towards the low-price promotion dimension; and if a competitor advertisement category frequently displays a diamond-embedded car logo or a star endorsement, the competitor advertisement category is biased towards a high luxury feature value.

[0111] In the implementation manner, the difference between the brand style feature value corresponding to the competitor video advertisement category and the target brand style feature value can be further calculated as the brand style difference value, and the brand style difference value can objectively reflect the matching degree of the brand positioning.

[0112] For example, when the target product emphasizes cost-effectiveness, a competitor advertisement category with a high luxury feature value will generate a larger positive difference value, revealing that the brand style of the competitor advertisement category with a high luxury feature value is significantly deviated from the target product, and thus the synergy of the advertisement strategy and the brand positioning can be evaluated.

[0113] S260, determine the sum of the advertisement performance evaluation value, the budget factor difference value, and the brand style difference value corresponding to the plurality of competitor video advertisement categories respectively as the advertisement category evaluation value. Determine the maximum advertisement category evaluation value in the advertisement category evaluation values corresponding to the plurality of competitor video advertisement categories respectively, and take the competitor video advertisement category corresponding to the maximum advertisement category evaluation value as the target competitor video advertisement category.

[0114] In the implementation manner, the sum of the advertisement performance evaluation value, the budget factor difference value, and the brand style difference value can be further calculated to obtain a comprehensive advertisement category evaluation value. The advertisement performance evaluation value is derived from the historical effect index, the budget factor difference value reflects the cost matching degree, and the brand style difference value measures the positioning consistency, so that the advertisement category evaluation value can comprehensively reflect the advertisement performance evaluation value, the budget factor difference value, and the brand style difference value.

[0115] In the determination of the target competitive product video advertisement category, by comparing the advertisement category evaluation values of all categories, the maximum value of the target competitive product video advertisement category can be selected as the target competitive product video advertisement category, so as to ensure that the target competitive product video advertisement category achieves the optimal balance in the three dimensions of propagation effect, cost control and brand positioning.

[0116] The beneficial effects brought by the above implementation manners are that the introduction of the brand style difference value can avoid the risk of deviation of creative design and brand core value, and can avoid strategic mismatch caused by misusing promotion style for high-end brands or imitating luxury route for mass brands; can cover three key dimensions of propagation performance, cost adaptation and brand style, and can generate creative solutions more in line with the business strategy of the enterprise, thereby improving the comprehensive benefits of advertisement delivery.

[0117] Figure 4 The third advertisement creative design method based on advertisement sample analysis provided by the embodiments of the present application is shown in the flowchart as shown in Figure 4 In the S120, the target competitive product video advertisement category in the plurality of competitive product video advertisement categories is determined, and the S120 further includes S310 to S320, which are described below.

[0118] In S310, the historical style evaluation value of the historical advertisement of the target product is obtained, and the style evaluation value of the competitive product video advertisement in the plurality of competitive product video advertisement categories is obtained. The difference between the style evaluation value of the competitive product video advertisement in the plurality of competitive product video advertisement categories and the historical style evaluation value is determined as the advertisement style innovation value. The style evaluation value includes a plurality of dimensional style evaluation feature values.

[0119] In the present implementation manner, in the advertisement creative design process, the historical style evaluation value of the historical advertisement of the target product can be obtained, and the historical style evaluation value represents the style features of the historical advertisement of the target product.

[0120] For example, the historical style evaluation value can be obtained by quantitative analysis of the multi-dimensional style features of the historical advertisement of the target product, and the historical style evaluation value can include evaluation feature values in the dimensions of picture color tone features, shot switching frequency and narrative rhythm.

[0121] In the present implementation manner, the style evaluation value of the competitive advertisement in the plurality of competitive product video advertisement categories can be obtained, and the style evaluation value reflects the characteristics of the advertisement strategies of different style competitive advertisements in creative performance.

[0122] For example, for beverage video advertisements, the historical advertisement of the target product may use bright color tone and high-frequency shot switching, and if a competitive advertisement category uses low-saturation cold color tone and long shot narrative, the historical style evaluation value and the style evaluation value present significant difference features.

[0123] S320, determine the sum of the advertisement performance evaluation value, the budget factor difference value, the brand style difference value and the advertisement style innovation value corresponding to each of the plurality of competitive video advertisement categories as the advertisement category evaluation value. Determine the maximum advertisement category evaluation value in the advertisement category evaluation values corresponding to each of the plurality of competitive video advertisement categories, and take the competitive video advertisement category corresponding to the maximum advertisement category evaluation value as the target competitive video advertisement category.

[0124] In the implementation manner, the difference between the style evaluation value corresponding to the competitive video advertisement category and the historical style evaluation value can be calculated as the advertisement style innovation value, and the advertisement style innovation value can quantitatively reflect the breakthrough degree of the inherent style of the advertisement strategy and the target product of the competitive video advertisement category.

[0125] For example, when a certain electronic product competitive advertisement category adopts a documentary interview form instead of a traditional product display, the difference forms a positive innovation value; if the same star endorsement mode is used, the innovation value tends to be zero, so that the advertisement style innovation value can dynamically evaluate the balance between advertisement innovation and brand continuity.

[0126] After obtaining the advertisement style innovation value, the advertisement performance evaluation value, the budget factor difference value, the brand style difference value and the advertisement style innovation value can be further superimposed to generate a more comprehensive advertisement category evaluation value; wherein the advertisement performance evaluation value measures the propagation efficiency, the budget factor difference value controls the cost risk, the brand style difference value ensures the strategic synergy, and the advertisement style innovation value stimulates the creative breakthrough of the advertisement.

[0127] After obtaining the advertisement category evaluation value, all advertisement category evaluation values can be compared, the competitive video advertisement category corresponding to the maximum advertisement category evaluation value is taken as the target competitive video advertisement category, and then the target competitive video advertisement category can be determined.

[0128] The implementation manner has the beneficial effects that the advertisement style innovation value index stimulates the selection of innovative strategies, can avoid the creative design from being trapped in the brand historical path dependence, improves the market freshness and topicality of the advertisement, can simultaneously consider the balanced development of the propagation effect, the budget adaptation, the brand matching and the innovation breakthrough, maximizes the breakthrough value of the advertisement creativity on the premise of ensuring the stability of the business strategy, and optimizes the long-term competitiveness of the advertisement.

[0129] In some implementation manners, in the S120, determining the target competitive video advertisement category in the plurality of competitive video advertisement categories further includes S330 to S340, which are specifically described as follows.

[0130] S330, obtain the negative evaluation rate corresponding to the competitive video advertisement in the plurality of competitive video advertisement categories.

[0131] In the present implementation, in the process of creative design, the negative evaluation rate of the competitor video advertisement in the multiple competitor video advertisement categories can be obtained, and the negative evaluation rate reflects the proportion of negative feedback of users in the process of advertisement dissemination.

[0132] For example, for a fast-moving consumer goods video advertisement, a certain competitor advertisement category may have 30% of the comments mentioning the keywords such as "fake" or "excessive packaging" due to exaggerated propaganda, and the negative evaluation rate is significantly higher than the industry average level. The negative evaluation rate index can reveal the potential public opinion risk of the advertisement strategy.

[0133] S340, determine the product of the advertisement category evaluation value and the negative evaluation rate corresponding to each of the multiple competitor video advertisement categories and normalize it as an advertisement category evaluation adjustment value. Determine the maximum advertisement category evaluation adjustment value in the advertisement category evaluation adjustment values corresponding to the multiple competitor video advertisement categories, and take the competitor video advertisement category corresponding to the maximum advertisement category evaluation adjustment value as the target competitor video advertisement category.

[0134] After obtaining the negative evaluation rate, the quotient value of the advertisement category evaluation value corresponding to each competitor video advertisement category and the negative evaluation rate can be calculated, and the calculation result can be normalized to obtain an advertisement category evaluation adjustment value; wherein the advertisement category evaluation value is derived from the sum of the advertisement performance evaluation value, the budget factor difference value, the brand style difference value and the advertisement style innovation value in the previous step, and the division operation will significantly reduce the advertisement category evaluation adjustment value of the competitor video advertisement category with high negative evaluation rate.

[0135] For example, a certain makeup video advertisement category has a high advertisement category evaluation value of 8.5 (creative and budget adaptation), but due to the negative evaluation rate of 18%, the normalized adjustment value is reduced to 7.2, so that the advertisement category evaluation adjustment value is significantly lower than other competitor categories with a negative rate of only 5%.

[0136] After obtaining the advertisement category evaluation adjustment value, by comparing the advertisement category evaluation adjustment values of all categories, the competitor video advertisement category corresponding to the maximum value of the advertisement category evaluation adjustment value can be selected as the target category, which can ensure that the selected scheme has both effect advantage and risk control ability.

[0137] The above-mentioned implementation has the beneficial effects that the comprehensive evaluation value is dynamically inhibited by the negative evaluation rate, which can effectively avoid the selection of high propagation risk advertisement strategy and prevent the damage of brand reputation; the public opinion monitoring index is included in the evaluation system, so that the finally selected advertisement category has stronger public opinion stability while maintaining creative effect, and the long-term sustainability of marketing activities is improved.

[0138] The embodiment of the present application also provides an advertisement creative design system based on advertisement sample analysis, which comprises units for executing the method as described in any of the above.

[0139] Figure 5 A first advertisement creative design system based on advertisement sample analysis provided by the embodiment of the present application has a logical structure diagram as shown in the figure. Figure 5 The system 1 of the embodiment comprises a processing unit 11, a storage unit 12 and a transceiver unit 13, the processing unit 11 is used for processing data, the storage unit 12 is used for storing data, and the transceiver unit 13 is used for transceiving data, and the processing unit 11, the storage unit 12 and the transceiver unit 13 cooperate with each other to realize the method described above. The beneficial effects of the embodiment of the present application have been described in the method, which will not be repeated here.

[0140] It should be noted that the information interaction, execution process and the like between the above-mentioned devices / units are based on the same concept as the method embodiments of the present application, and the specific functions and the technical effects brought by them can be referred to the method embodiments part, which will not be repeated here.

[0141] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit, and the integrated unit can be realized in the form of hardware or software. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the units and modules in the system can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0142] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the present application can implement all or part of the processes in the above-mentioned embodiment methods through a computer program to instruct relevant hardware to complete, and the computer program can be stored in a computer readable storage medium. When the computer program is executed by a processor, the steps of each method embodiment described above can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms. The computer readable medium at least includes any entity or device capable of carrying the computer program code to the photographing device / terminal equipment, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium. For example, U disk, mobile hard disk, magnetic disk or optical disk, etc. In some jurisdictions, according to legislation and patent practice, the computer readable medium can not be an electrical carrier signal and a telecommunication signal.

[0143] In the above embodiments, the description of each embodiment has its own focus, and the parts not described or recorded in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0144] Those skilled in the art can appreciate that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether the functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. The skilled person can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0145] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division, and actual implementation can have another division manner. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0146] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may also be distributed to multiple network units. Part or all of the units can be selected to achieve the purpose of the embodiment scheme according to actual needs.

[0147] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can still be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for advertising creative design based on advertising sample analysis, characterized in that, The method includes: Obtain multiple competitor video ads for the target product, and obtain keyframe screenshots, positive comment times, and positive comment keyword clouds for high completion rate segments corresponding to each competitor video ad; Based on keyframe screenshots, positive comment times, and positive comment keyword clouds of high completion rate segments corresponding to multiple competitor video ads, multiple competitor video ads are clustered to obtain multiple competitor video ad categories, and the target competitor video ad category is determined among multiple competitor video ad categories; Using an advertising creative design model, we determine the corresponding advertising creative design scheme for the target product based on multiple competitor video ads in the target competitor video ad category.

2. The method as described in claim 1, characterized in that, Based on keyframe screenshots, positive comment times, and positive comment keyword clouds of high-completion-rate segments corresponding to multiple competitor video ads, clustering is performed on the multiple competitor video ads to obtain multiple competitor video ad categories, including: By using a keyframe recognition model corresponding to the target product, and based on keyframe screenshots of high-completion-rate segments of multiple competitor video ads, the features of each ad segment are determined. These ad segment features represent the image content features of the keyframe screenshots of high-completion-rate segments. Furthermore, by using a causal relationship derivation model, and based on the timestamps, positive comment times, and positive comment keyword cloud corresponding to the keyframe screenshots of high-completion-rate segments, causal relationship features are determined. These causal relationship features represent the causal relationship between the keyframe screenshots of high-completion-rate segments and the positive comment times and positive comment keyword cloud. Based on the ad segment features and causal relationship features corresponding to multiple competitor video ads, multiple competitor video ads are clustered to obtain multiple competitor video ad categories.

3. The method as described in claim 2, characterized in that, Based on keyframe screenshots, positive comment times, and positive comment keyword clouds of high-completion-rate segments corresponding to multiple competitor video ads, clustering is performed on the multiple competitor video ad categories, including: Determine the product similarity between the target product and the competing products corresponding to multiple competitor video ads; Based on the product similarity, ad segment features, and causal relationship features corresponding to multiple competitor video ads, multiple competitor video ads are clustered to obtain multiple competitor video ad categories.

4. The method as described in claim 3, characterized in that, Based on keyframe screenshots, positive comment times, and positive comment keyword clouds of high-completion-rate segments corresponding to multiple competitor video ads, clustering is performed on the multiple competitor video ad categories, including: Determine the similarity of core selling points, packaging methods, and advertising strategies of the competing products and the target product corresponding to multiple competing product video ads. Then, determine the sum of the reciprocals of the similarity of core selling points, packaging methods, and advertising strategies of the competing products corresponding to multiple competing product video ads, and use this sum as the difference in competitive factors corresponding to multiple competing product video ads. Based on the differences in competitive factors, ad segment characteristics, and causal relationship characteristics of multiple competitor video ads, clustering is performed on multiple competitor video ads to obtain multiple competitor video ad categories.

5. The method as described in claim 4, characterized in that, Identify the target competitor video ad category from among multiple competitor video ad categories, including: Obtain user comment density, positive comment sentiment index, and completion rate of competitor video ads in multiple competitor video ad categories within a preset time period; based on user comment density, positive comment sentiment index, and completion rate of competitor video ads in multiple competitor video ad categories within a preset time period, determine the ad performance evaluation value corresponding to each of the multiple competitor video ad categories; Determine the highest ad performance rating among the ad performance ratings for multiple competitor video ad categories, and use the competitor video ad category corresponding to the highest ad performance rating as the target competitor video ad category.

6. The method as described in claim 5, characterized in that, Identifying the target competitor video ad category among multiple competitor video ad categories also includes: Obtain the target budget factor corresponding to the target product, and obtain the budget factors corresponding to competitor video ads in multiple competitor video ad categories; determine the difference between the budget factors corresponding to competitor video ads in multiple competitor video ad categories and the target budget factor, as the budget factor difference value; where the budget factor represents the budget amount of the video ad; The sum of the ad performance evaluation values ​​and budget factor differences for each of the multiple competitor video ad categories is determined as the ad category evaluation value. The maximum ad category evaluation value among the multiple competitor video ad categories is determined, and the competitor video ad category corresponding to the maximum ad category evaluation value is taken as the target competitor video ad category.

7. The method as described in claim 6, characterized in that, Identifying the target competitor video ad category among multiple competitor video ad categories also includes: Obtain the target brand style feature value corresponding to the target product, and obtain the brand style feature value corresponding to the competitor video ads in multiple competitor video ad categories; determine the difference between the brand style feature value corresponding to the competitor video ads in multiple competitor video ad categories and the target brand style feature value, as the brand style difference value; whereby the brand style feature value represents the low-price promotion feature and luxury feature of the video ad. The sum of the ad performance evaluation value, budget factor difference value, and brand style difference value corresponding to multiple competitor video ad categories is determined as the ad category evaluation value; the maximum ad category evaluation value among the ad category evaluation values ​​corresponding to multiple competitor video ad categories is determined, and the competitor video ad category corresponding to the maximum ad category evaluation value is taken as the target competitor video ad category.

8. The method as described in claim 7, characterized in that, Identifying the target competitor video ad category among multiple competitor video ad categories also includes: Obtain the historical style evaluation value of the target product's historical advertisements, and obtain the style evaluation value of competitor video advertisements in multiple competitor video advertisement categories; determine the difference between the style evaluation value and the historical style evaluation value of competitor video advertisements in multiple competitor video advertisement categories as the advertisement style innovation value; wherein, the style evaluation value includes style evaluation feature values ​​of multiple dimensions; The sum of the ad performance evaluation value, budget factor difference value, brand style difference value, and ad style innovation value corresponding to multiple competitor video ad categories is determined as the ad category evaluation value; the maximum ad category evaluation value among the ad category evaluation values ​​corresponding to multiple competitor video ad categories is determined, and the competitor video ad category corresponding to the maximum ad category evaluation value is taken as the target competitor video ad category.

9. The method as described in claim 8, characterized in that, Identifying the target competitor video ad category among multiple competitor video ad categories also includes: Obtain the negative review rate of competitor video ads across multiple competitor video ad categories; Determine the quotient and normalize the ad category evaluation value and negative evaluation rate corresponding to each of the multiple competitor video ad categories, and use them as the ad category evaluation adjustment value; determine the maximum ad category evaluation adjustment value among the ad category evaluation adjustment values ​​corresponding to each of the multiple competitor video ad categories, and take the competitor video ad category corresponding to the maximum ad category evaluation adjustment value as the target competitor video ad category.

10. An advertising creative design system based on advertising sample analysis, characterized in that, Includes a unit for performing the method according to any one of claims 1 to 9.

Images