Brand AI cognitive level evaluation method and device, equipment and storage medium

Abstract

The invention provides a brand AI cognitive level assessment method and device, equipment and a storage medium. The method comprises the following steps: acquiring an industry to which a to-be-assessed brand belongs and a brand name; generating a plurality of keywords of the brand to be evaluated based on the industry and the brand name; aiming at each keyword, respectively constructing a plurality of Prompt questions; a plurality of AI large models are adopted to answer the Prompt questions, and answers of the Prompt questions are obtained; and based on the answers, generating a cognitive level assessment result of the brand to be assessed in the AI ecology. By implementing the method provided by the invention, the cognitive level evaluation of the brand in the AI ecology can be accurately and effectively realized, so that a solid foundation is laid for improving the brand influence of the brand in the AI ecology.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, and more particularly to a method, apparatus, device, and storage medium for assessing a brand's AI cognitive level. Background Technology

[0002] With the rise of large-scale AI models and their emergence as an important tool in people's lives and work, more and more people are using AI tools to search for answers, such as which car to buy or which shampoo to use. Therefore, the AI's responses are crucial, as its recommendation biases directly impact a brand's market performance and consumer perception. However, AI responses are inherently constrained by its internal level of brand awareness. Therefore, for brands, accurately and effectively assessing their brand awareness level within AI and then implementing targeted strategies to enhance brand influence in the AI ​​ecosystem has become a pressing technical challenge for those skilled in the art. Summary of the Invention

[0003] In view of this, this disclosure proposes a method, apparatus, device and storage medium for assessing a brand’s AI cognitive level, which can accurately and effectively assess a brand’s cognitive level in the AI ​​ecosystem.

[0004] According to a first aspect of this disclosure, a method for assessing a brand's AI awareness level is provided, comprising: Obtain the industry and brand name of the brand to be evaluated; Based on the industry and the brand name, generate multiple keywords for the brand to be evaluated; For each of the aforementioned keywords, construct multiple Prompt questions; Multiple large AI models are used to answer each of the aforementioned Prompt questions, thus obtaining the answers to each of the aforementioned Prompt questions; Based on the answers provided, an assessment result is generated regarding the brand's cognitive level within the AI ​​ecosystem.

[0005] In one possible implementation, when generating multiple keywords for the brand to be evaluated based on the industry and the brand name, the process includes: Obtain pre-built keyword prompt templates; The industry and brand name are injected into the keyword generation prompt template to obtain keyword generation prompts; Based on the keywords, prompt words are generated, and the AI ​​model is used to generate multiple keywords for the brand to be evaluated.

[0006] In one possible implementation, when constructing multiple Prompt questions for each of the aforementioned keywords, the following is included: Obtain a pre-built question prompt template; The industry and each of the keywords are injected into the question generation prompt template to obtain question generation prompts; Based on the question-generating prompts and the AI ​​big model, multiple prompt questions are constructed for each of the keywords.

[0007] In one possible implementation, when generating the assessment result of the brand's cognitive level in the AI ​​ecosystem based on the answers provided, the process includes: The corresponding answer information is extracted from each of the answers, wherein the answer information includes at least one of the recommended brand name, search article information, and cited article information; Based on the answer information corresponding to each of the above answers, an assessment result of the brand's cognitive level in the AI ​​ecosystem is generated.

[0008] In one possible implementation, when generating the assessment result of the brand's cognitive level in the AI ​​ecosystem based on the answer information corresponding to each of the answers, the process includes: Based on the answer information corresponding to each of the answers, the index value of the preset cognitive level index is calculated, wherein the cognitive level index includes at least one of brand visibility, citation rate of brand promotion articles, and citation rate of the main text. Based on the index values ​​of the cognitive level indicators, an assessment result of the cognitive level of the brand to be evaluated in the AI ​​ecosystem is generated.

[0009] In one possible implementation, after generating the assessment results of the brand's cognitive level in the AI ​​ecosystem, the method further includes: The results of the cognitive level assessment are then visualized.

[0010] In one possible implementation, after generating the assessment results of the brand's cognitive level in the AI ​​ecosystem, the method further includes: Based on the assessment results of the brand's awareness level in the AI ​​ecosystem, the brand's promotional strategy in the AI ​​ecosystem will be adjusted.

[0011] According to a second aspect of this disclosure, a device for assessing a brand's AI cognitive level is provided, comprising: The data acquisition module is used to obtain the industry and brand name of the brand to be evaluated; The keyword generation module is used to generate multiple keywords for the brand to be evaluated based on the industry and the brand name; The question generation module is used to generate multiple Prompt questions for each of the aforementioned keywords. The answer module is used to answer each of the Prompt questions using multiple large AI models, and obtain the answers to each of the Prompt questions; The evaluation module is used to generate an evaluation result of the brand's cognitive level in the AI ​​ecosystem based on each of the answers.

[0012] According to a third aspect of this disclosure, a brand AI cognitive level assessment device is provided, comprising: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to perform the method described in the first aspect of this disclosure.

[0013] According to a fourth aspect of this disclosure, a non-volatile computer-readable storage medium is provided that stores computer program instructions thereon, wherein the computer program instructions, when executed by a processor, implement the method described in the first aspect of this disclosure.

[0014] This disclosure provides a method, apparatus, device, and storage medium for assessing a brand's AI awareness level. The method includes: obtaining the industry and brand name of the brand to be assessed; generating multiple keywords for the brand based on the industry and brand name; constructing multiple prompt questions for each keyword; using multiple large AI models to answer each prompt question to obtain the answers to each prompt question; and generating an assessment result of the brand's awareness level in the AI ​​ecosystem based on the answers. By implementing the method provided in this disclosure, the brand's awareness level in the AI ​​ecosystem can be accurately and effectively assessed, thereby laying a solid foundation for enhancing the brand's influence in the AI ​​ecosystem.

[0015] Other features and aspects of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0016] The accompanying drawings, which are included in and form part of this specification, illustrate exemplary embodiments, features, and aspects of this disclosure together with the specification and serve to explain the principles of this disclosure.

[0017] Figure 1 A flowchart illustrating a method for assessing a brand's AI awareness level according to an embodiment of this disclosure is shown. Figure 2 A flowchart illustrating a method for assessing a brand's AI awareness level according to another embodiment of this disclosure; Figure 3 A diagram illustrating the response output by an AI large model according to an embodiment of the present disclosure; Figure 4 A diagram showing the homepage interface of an AI cognitive level assessment system according to an embodiment of the present disclosure is provided. Figure 5 A platform page interface diagram of an AI cognitive level assessment system according to an embodiment of the present disclosure is shown. Figure 6 A diagram showing the source analysis page interface of an AI cognitive level assessment system according to an embodiment of the present disclosure is provided. Figure 7 A schematic block diagram of a brand AI cognition level assessment device according to another embodiment of the present disclosure is shown; Figure 8 A schematic block diagram of a brand AI cognition level assessment device according to an embodiment of the present disclosure is shown. Detailed Implementation

[0018] Various exemplary embodiments, features, and aspects of this disclosure will now be described in detail with reference to the accompanying drawings. The same reference numerals in the drawings denote elements that have the same or similar functions. Although various aspects of the embodiments are shown in the drawings, they are not necessarily drawn to scale unless specifically indicated otherwise.

[0019] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments.

[0020] Furthermore, to better illustrate this disclosure, numerous specific details are set forth in the following detailed description. Those skilled in the art will understand that this disclosure can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of this disclosure.

[0021] <Method Implementation> Figure 1 A flowchart illustrating a method for assessing a brand's AI awareness level according to an embodiment of this disclosure is shown. Figure 2 This document presents an example flowchart of a brand AI awareness assessment method according to an embodiment of the present disclosure, which is described below in conjunction with... Figure 1 and Figure 2 The method disclosed herein will be described in detail. For example... Figure 1 As shown, the method includes steps S1100-S1500.

[0022] S1100: Obtain the industry and brand name of the brand to be evaluated.

[0023] S1200 generates multiple keywords for the brand to be evaluated based on its industry and brand name. These keywords can include at least one of scenario-based keywords and brand positioning keywords. For example, for a car brand... The generated keywords can include: new energy vehicles, cars for young people, and cars under 300,000 yuan.

[0024] In one possible implementation, generating multiple keywords for a brand to be evaluated based on its industry and brand name may include the following steps: First, obtain a pre-built keyword generation prompt template. In a specific embodiment, this keyword generation prompt template may be as follows: Based on the (industry) and (brand name), generate keywords, which may be scenario keywords and brand positioning keywords. Next, inject the industry and brand name of the brand to be evaluated into the aforementioned keyword generation prompt template to obtain keyword generation prompts. Finally, based on the keyword generation prompts and the AI ​​model, generate multiple keywords for the brand to be evaluated. Specifically, input the generated keyword generation prompts into the AI ​​model to obtain the multiple keywords for the brand to be evaluated returned by the AI ​​model (i.e., Figure 2 (The brand keywords shown).

[0025] S1300: For each keyword generated in step S1200, construct multiple Prompt questions. This may include the following steps: First, obtain a pre-constructed question generation prompt template; second, inject the industry and each keyword into the prompt template to obtain the prompts; finally, based on the prompts and the AI ​​model, construct multiple Prompt questions for each keyword. Specifically, input the prompts into the AI ​​model to obtain multiple Prompt questions for each keyword returned by the AI ​​model (specifically as follows...). Figure 2 (As shown).

[0026] In one possible implementation, the question generation prompt template includes industry variables, keyword variables, and question generation instructions. The instructions further specify the requirements for question generation and the returned data format. These requirements may include at least one of the following: mimicking the tone of real consumers, diversifying questioning angles, eliminating brand name interference, and avoiding question duplication. Setting these requirements ensures the diversity and authenticity of generated questions, fully stimulating the cognitive capabilities of the AI ​​model while avoiding evaluation bias. In a specific example, the question generation prompt template might look like this: "Based on the brand's industry and keywords, generate prompt statements for the AI ​​model to ask questions. Each keyword must generate three different recommendation questions, with the following requirements: 1. Completely mimic the tone and questioning style of real consumers; 2. Questions should be diverse in perspective, including functional inquiries, usage scenarios, comparative consultations, etc.; 3. Avoid directly mentioning any brand name; 4. Questions should be naturally and fluently worded, avoiding repetition. My industry is (industry), and the keywords are (multiple keywords generated in step S1200). The generated prompt question list should be returned in JSON format, with the JSON format being [{"keyword":"keyword1","questions":["prompt question1","prompt question2",……"prompt question n"]},"keyword2",……]".

[0027] Based on this generated prompt template, the system can automatically generate multi-dimensional and multi-faceted consumer-style questions for each keyword. For example, for the keyword "moisturizing" in the skincare industry, possible prompt questions might include: 1. My skin has been particularly dry lately, can you recommend any good moisturizing products? 2. I have sensitive skin and would like some gentle moisturizing skincare products, any suggestions? 3. With autumn and winter approaching, which type of moisturizing product has a longer-lasting effect? ​​etc.

[0028] By generating multiple prompt questions for each keyword using the aforementioned prompt templates, the standardization and comprehensiveness of the questions during the evaluation process are ensured. Specifically, mimicking the tone of real consumers makes the generated prompt questions more relevant to actual usage scenarios; diverse question angles cover different cognitive dimensions such as function, scenario, and comparison; eliminating brand name interference avoids leading bias; and avoiding question repetition ensures a rich sample of questions. These requirements work together to more effectively stimulate the AI ​​model to demonstrate a complete picture of its brand perception, providing rich and diverse answer samples for subsequent cognitive level assessments, thereby significantly improving the accuracy and reliability of the evaluation results.

[0029] In one possible implementation, to improve the relevance and evaluation efficiency of the Prompt questions, after constructing multiple Prompt questions, the process includes: real-time analysis of the AI ​​model's response quality, and dynamic adjustment of the Prompt based on the response quality to ensure comprehensive coverage of brand awareness dimensions. Specifically, this may include the following steps: First, after the initial Prompt is generated, multiple large AI models are used to answer each Prompt question, and the results are collected. Simultaneously, a quality assessment module is built. This module analyzes each answer in real time based on preset indicators (such as answer relevance, frequency of brand keywords, and diversity score), generating a corresponding quality assessment score. The formula for calculating the quality assessment score for each answer is as follows: Q = α × R + β × D + γ × B Where Q is the quality assessment score, R is the answer relevance score (calculated based on semantic matching), D is the diversity score (calculated based on the number of cognitive dimensions covered in the answer), B is the frequency of brand keywords, and α, β, γ are preset weight coefficients, which can be dynamically adjusted according to industry characteristics.

[0030] Secondly, based on the quality assessment scores of each answer, the Prompt questions and keywords corresponding to low-quality answers are identified. For example, if the average quality score of Prompt questions for a certain keyword is lower than a preset threshold (such as 0.7), the Prompt optimization process is triggered.

[0031] Next, for the Prompt questions and keywords corresponding to low-quality answers, new Prompt questions are generated using an AI big data model, focusing on weak cognitive dimensions (such as adding functional comparisons, refining usage scenarios, etc.) and adjusting the wording to mimic the tone of real consumers in order to eliminate evaluation bias.

[0032] The above process is repeated iteratively until the quality assessment scores of all Prompt questions reach a stable state or the iteration termination condition is met (such as the maximum number of iterations or score convergence). Finally, the optimized Prompt question set is output for subsequent cognitive level assessment.

[0033] The above-mentioned Prompt optimization scheme can effectively solve the problem of incomplete evaluation caused by static Prompt templates, and improve the accuracy and efficiency of evaluation results.

[0034] S1400 employs multiple large AI models to answer each Prompt question, thus obtaining the answers to each Prompt question. These multiple large AI models originate from multiple different platforms; for example, the multiple large AI models could be from different platforms. , Bao, Wenxin words, ki i and DEEPSE K, etc. Each answer includes a text description and / or image of the recommended brand, links to articles found during the AI's response process (referred to as search articles), and links to articles cited when generating the AI ​​answer (referred to as cited articles). In a specific example, one of the large AI models returned an answer such as... Figure 3 As shown.

[0035] In one possible implementation, when using multiple large AI models to answer each Prompt question and obtaining the answers to each Prompt question, the following steps may be included: input each Prompt question into each large AI model respectively, and then obtain the answers to each Prompt question returned by each large AI model. That is, each Prompt question will be answered by multiple large AI models to achieve cross-validation, obtain a comprehensive cognitive picture, and avoid the bias of a single model.

[0036] To further eliminate data bias caused by regional server preferences and user account characteristics, a distributed, multi-factor data collection model will be adopted. Specifically, independent computing nodes will be deployed in multiple different cities, with an identical set of large AI models deployed on each node. Through remote control technology, real user behavior will be simulated, with different accounts taking turns submitting prompts to the various AI models on computing nodes in different cities. This design ensures that each prompt collects responses from different regions, different access accounts, and different AI models, thereby fundamentally guaranteeing the diversity of data sources and the objectivity of evaluation conclusions.

[0037] S1500, based on the answers obtained in step S1400, generates an assessment result of the brand's cognitive level in the AI ​​ecosystem. This may specifically include the following steps: First, extract the corresponding answer information from each answer. This answer information includes at least one of the following: the recommended brand name, search article information, and cited article information. The search article information includes at least one of the following: the name of the search article, the brand keywords it covers, the publishing website, the article link, and the publication time. The cited article information includes at least one of the following: the name of the cited article, the brand keywords it covers, the publishing website, the article link, and the publication time.

[0038] Specifically, for each answer received: First, pre-built brand extraction prompts are obtained. These prompts, along with the answer itself, are input into an AI model, which then extracts the recommended brand names mentioned in the answer. In a specific example, the pre-built brand extraction prompts might look like this: Based on the text description and / or images in the answer, all the recommended brand names appearing in sequence, separated by commas. Second, links to the searched articles and cited articles are identified from the answer. This information is then extracted using web scraping techniques such as Python and PlayWright. Following this method, the corresponding answer information is extracted from each answer.

[0039] In one possible implementation, after obtaining the answer information corresponding to each answer, all answer information for the current round will be summarized according to the collection date, the brand name to be evaluated, keywords, prompt question, the AI ​​model used, and the form of answer information to obtain the brand awareness analysis dataset for the current round, so as to facilitate subsequent data statistical analysis.

[0040] In one possible implementation, after obtaining the brand awareness analysis dataset for the current round, data cleaning and standardization are performed on each record of the brand awareness analysis dataset to obtain a high-quality, standardized brand awareness analysis dataset.

[0041] Second, based on the answer information corresponding to each answer, an assessment result of the brand's cognitive level in the AI ​​ecosystem is generated.

[0042] In one possible implementation, when generating an assessment result of the brand's cognitive level in the AI ​​ecosystem based on the answer information corresponding to each answer, the following steps may be included: First, based on the answer information corresponding to each answer, calculate the index value of the preset cognitive level indicator, wherein the cognitive level indicator includes at least one of brand visibility, citation rate of brand promotion articles, and citation rate of the main text.

[0043] Brand visibility is used to analyze whether a brand can reach consumers through AI big data models. Higher brand visibility makes it easier for the brand to reach consumers via AI big data models. Brand visibility is further subdivided into overall brand visibility across all platforms, overall brand visibility across all platforms, brand keyword visibility across all platforms, and brand keyword visibility across all platforms. Through these subdivided brand visibility metrics, the strengths and weaknesses of brand awareness can be accurately identified, providing data support for developing differentiated, precise, and efficient advertising strategies. The calculation formulas for the various subdivided brand visibility metrics are shown below: In the formula, To improve the brand's overall visibility across all platforms, This refers to the number of times the brand name was included in all responses generated during the current round of evaluation. The number of all prompts initiated for the current round of evaluation.

[0044] In the formula, To improve the brand's overall visibility across various platforms, This refers to the number of times the brand name was included in all responses generated by each platform during the current round of evaluation. This represents the number of Prompt questions initiated by each platform during the current round of evaluation.

[0045] In the formula, To improve the visibility of each brand's keywords across all platforms, This refers to the number of times the brand name is included in all responses generated for each keyword during the current evaluation round. This represents the number of prompt questions initiated for each keyword during the current round of evaluation.

[0046] In the formula, To measure the visibility of each brand's keywords across various platforms, This refers to the number of times the brand name is included in all responses generated across various platforms for each keyword during the current evaluation round. This represents the number of Prompt questions initiated on each platform for each keyword during the current round of evaluation.

[0047] The citation rate of brand promotion articles is used to analyze the probability that articles published by the brand will be searched by the AI ​​model and included in the candidate knowledge base as information sources. The higher the citation rate, the higher the value of the brand promotion article, and the more important it is to maintain and optimize it. The citation rate of brand promotion articles is calculated as follows: Citation rate of brand promotion articles = Total number of times the brand promotion article is searched by the AI ​​model in the current round of evaluation / Total number of Prompt questions initiated in the current round of evaluation.

[0048] The citation rate of the promotional article is used to analyze the probability that the brand's promotional copy is adopted by the AI ​​model as an information source and directly used to support the core arguments of the answer. The higher the citation rate, the higher the effectiveness of the promotional article content, thus helping the brand understand which content is valuable and will be adopted by the AI. The formula for calculating the citation rate of the promotional article is as follows: Citation Rate of Promotional Article = Total Number of Times the Article is Annotated in the Text Among All Answers Generated in the Current Evaluation Round / Number of Prompt Questions Asked in the Current Evaluation Round.

[0049] Second, based on the index values ​​of cognitive level indicators, an assessment result of the cognitive level of the brand to be evaluated in the AI ​​ecosystem is generated. Specifically, the visibility of each keyword for each brand across the entire platform is aggregated, and each brand is ranked accordingly to obtain a brand visibility ranking. The citation rate and text citation rate of the brand's promotional articles are aggregated into the corresponding search article information or cited article information, and each search article and cited article is grouped according to the brand keywords covered. Within each group, they are sorted according to citation rate or text citation rate to obtain the source information of the brand to be evaluated. The calculated brand visibility of each segment, brand visibility ranking, and source information of the brand to be evaluated are used as the evaluation result for that brand.

[0050] In one possible implementation, after generating the cognitive level assessment results for the brand to be evaluated within the AI ​​ecosystem, the process also includes visualizing these results. Specifically, the cognitive level assessment results for the brand to be evaluated are displayed in a paginated format.

[0051] On the homepage (i.e.) Figure 4 The Home page shown displays the overall visibility of the brand to be evaluated across all platforms and the visibility of each keyword across all platforms (e.g., ...). Figure 4 The content shown in the Brand Visibility section) and the brand's overall visibility across various platforms (such as...) Figure 4 (The content shown in the Models section). In this way, you can quickly grasp the overall perception performance of the brand to be evaluated in the AI ​​ecosystem through the homepage, identify the advantageous platforms and keyword dimensions that need to be improved, and establish a macro overview of the brand's AI perception level.

[0052] On the platform page (i.e.) Figure 5 The Model page shown displays various keywords for the brand (such as...). Figure 5 The BrandVisibility section displays the content shown therein. Clicking on different keywords will show the visibility of the selected keyword on various platforms (e.g., ...). Figure 5The content displayed in the "Brand Available by AI Model" section of the page will also show the overall visibility ranking of different brands across all platforms for the selected keywords (e.g., ...). Figure 5 (The content shown in the Brand IndustryRanking section). In this way, the competitive landscape under specific keywords can be analyzed in depth through the platform page, accurately identifying the brand's cognitive weaknesses and strengths in specific segmented scenarios, and identifying the differences in the cognitive understanding of specific keywords among different AI platforms, providing a basis for formulating differentiated platform operation strategies.

[0053] On the source analysis page (i.e.) Figure 6 The Citation page shown displays source information for the brand to be evaluated (such as...). Figure 6 The content shown in the Details section. Simultaneously, the source information is analyzed to determine the distribution of each article's publishing platform, and the source information page (such as...) is displayed. Figure 6 (The content shown in the Citation section). In this way, the source analysis page can trace the root sources of content that influence AI's cognition, identify high-value articles and their publishing platforms that are frequently searched and cited by the AI ​​model, and thus provide clear data guidance for optimizing content strategies and improving the quality and influence of sources.

[0054] In one possible implementation, after generating an assessment of the brand's cognitive level within the AI ​​ecosystem, the brand's promotional strategy can be systematically adjusted based on these results. The specific adjustment process can employ a data-driven closed-loop optimization approach as follows: First, by using the overall visibility data displayed on the homepage, we identify target platforms where the overall visibility of the brand to be evaluated is lower than the preset first threshold across all platforms, and these platforms are the ones that need to be prioritized for optimization.

[0055] Subsequently, for each target platform, we further analyzed its keyword performance on the platform page. By examining the visibility data of each keyword on the platform, we filtered out target keywords with visibility below a preset second threshold, thereby accurately locating the specific scenarios or sub-sectors where the cognitive gap lies.

[0056] Furthermore, based on the target keywords, the visibility ranking of each brand under the target keywords is obtained through the brand industry ranking function provided on the platform page, thereby clarifying the competitive position of the brand to be evaluated and identifying the competitors that perform well under the target keywords.

[0057] Building upon this, the system navigates to the information source analysis page, where it filters out brand promotional articles published by the brand to be evaluated and its competitors targeting the same keyword. By comparing the characteristics of both parties' articles across different platforms, content themes, and writing structures, the system identifies key factors leading to differences in perception, thereby generating targeted adjustments to the promotional strategy.

[0058] For example, if analysis reveals that competitors' promotional articles targeting the target keywords are highly concentrated on website A, and the content on that website is frequently cited by the AI ​​model, while the brand to be evaluated lacks corresponding content on website A, then a strategy can be developed: for the target keywords, referencing the content angles and styles of competitors, generate diverse brand promotional copy, and publish and promote it centrally on website A to quickly fill the content gap and improve AI visibility.

[0059] For example, if a brand to be evaluated has published content on website A, but its articles are not effectively searched or cited in AI responses, it is necessary to conduct in-depth analysis of the content strategies of competitors' highly cited articles, such as their keyword layout, information presentation format, or content depth, and optimize the brand's published articles accordingly to increase the probability of them being recognized and adopted by AI.

[0060] Through the above-mentioned step-by-step drilling and analysis, from macro-level platform positioning to micro-level keyword and content diagnosis, this method can provide precise and actionable decision support for brands to adjust their promotional strategies in the AI ​​ecosystem, thereby achieving continuous improvement in cognitive level.

[0061] In one possible implementation, after generating the assessment results of the brand's cognitive level in the AI ​​ecosystem, the process further includes: automatically generating optimized articles based on these assessment results to help the brand adjust its promotional strategies within the AI ​​ecosystem. The specific generation method includes the following steps: First, the target keywords are determined using the method described above. Then, for each target keyword, the top three brands ranking under that keyword and their corresponding frequently cited articles are identified. Next, the article framework and narrative style of these frequently cited articles are summarized using an AI big data model. Finally, the brand name, keywords, and core viewpoints of the brand to be assessed are input into the AI ​​big data model, allowing it to generate new optimized article content based on the input content and mimicking the summarized article framework and narrative style of the frequently cited articles. In this way, brands can quickly generate high-quality promotional content that aligns with AI's cognitive patterns, effectively enhancing their brand influence within the AI ​​ecosystem.

[0062] In another possible implementation, a customized advertising strategy for the brand to be evaluated can be automatically generated based on the results of the cognitive level assessment. This may include the following steps: First, based on the cognitive level assessment results, the cognitive weaknesses of the brand to be evaluated under the target platforms and target keywords are identified. Specifically, through visualized data from the homepage and platform pages, platforms and keyword combinations with brand visibility below a preset threshold are located, forming a set of "platform-keyword" pairs to be optimized.

[0063] Next, for each "platform-keyword" pair, competitive intelligence analysis is performed: The top N competing brands in terms of visibility under that keyword are retrieved from the platform page, and frequently cited articles corresponding to these competing brands are extracted through source analysis. Subsequently, a strategy analysis prompt template is constructed, injecting key information from the frequently cited articles of the competing brands—including article titles, core viewpoints, content framework, narrative style, and publishing platforms—as input variables to generate strategy analysis prompts. These prompts are then input into an AI model, which automatically extracts the content strategy characteristics and success factors of the competing brands under that keyword.

[0064] Then, based on the extracted competitive intelligence, a strategy generation prompt template is constructed. The brand name, target keywords, core brand information, and content strategy characteristics and success factors of competing brands under the same keywords are injected into this template to generate strategy generation prompts. These prompts are then input into an AI model, which automatically generates a customized promotional strategy draft for the "platform-keyword" pair, including specific content direction, suggested publishing platforms, and optimization points.

[0065] Finally, the generated draft promotional strategy undergoes manual review and necessary revisions to form an executable final promotional strategy plan, upon which content creation and distribution are carried out. The system will record the correspondence between the strategy and the execution results for subsequent evaluation and iteration, thereby achieving a complete closed loop from cognitive assessment to strategy generation, content optimization, and effect tracking.

[0066] By introducing the aforementioned method for automatically generating promotional strategies, this disclosure can significantly improve the efficiency and accuracy of brands in optimizing their promotional strategies within the AI ​​ecosystem, achieving an intelligent leap from "diagnosing problems" to "prescribing remedies," and providing continuous and efficient decision support for enhancing the brand's influence in the AI ​​era.

[0067] This disclosure provides a method for assessing a brand's AI awareness level, including: obtaining the industry and brand name of the brand to be assessed; generating multiple keywords for the brand based on the industry and brand name; constructing multiple prompt questions for each keyword; using multiple large AI models to answer each prompt question, obtaining answers to each prompt question; and generating an assessment result of the brand's awareness level in the AI ​​ecosystem based on each answer. By implementing the method provided in this disclosure, the brand's awareness level in the AI ​​ecosystem can be accurately and effectively assessed, thereby laying a solid foundation for enhancing the brand's influence in the AI ​​ecosystem.

[0068] <Device Embodiment> Figure 7 A schematic block diagram of a brand AI cognition level assessment device according to an embodiment of the present disclosure is shown. Figure 7 As shown, the device 100 includes: The data acquisition module 110 is used to acquire the industry and brand name of the brand to be evaluated; The keyword generation module 120 is used to generate multiple keywords for the brand to be evaluated based on the industry and the brand name; The question generation module 130 is used to construct multiple Prompt questions for each of the aforementioned keywords; The answer module 140 is used to answer each of the Prompt questions using multiple large AI models to obtain the answers to each of the Prompt questions; The assessment module 150 is used to generate an assessment result of the brand's cognitive level in the AI ​​ecosystem based on each of the answers.

[0069] <Equipment Example> Figure 8 A schematic block diagram of a brand AI awareness assessment device according to an embodiment of the present disclosure is shown. Figure 8 As shown, the brand AI cognitive level assessment device 200 includes a processor 210 and a memory 220 for storing executable instructions of the processor 210. The processor 210 is configured to implement any of the aforementioned brand AI cognitive level assessment methods when executing the executable instructions.

[0070] It should be noted here that the number of processors 210 can be one or more. Furthermore, the AI ​​cognitive level assessment device 200 of this embodiment may also include an input device 230 and an output device 240. The processors 210, memory 220, input device 230, and output device 240 can be linked via a bus or other means, without specific limitations here.

[0071] The memory 220, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and various modules, such as the program or module corresponding to the brand's AI cognitive level assessment method in this embodiment of the disclosure. The processor 210 executes various functional applications and data processing of the brand's AI cognitive level assessment device 200 by running the software program or module stored in the memory 220.

[0072] Input device 230 can be used to receive input digital numbers or signals. These signals may include key signals related to user settings and function control of the device / terminal / server. Output device 240 may include a display device such as a screen.

[0073] <Storage Medium Examples> According to a fourth aspect of this disclosure, a non-volatile computer-readable storage medium is also provided, having stored thereon computer program instructions that, when executed by processor 210, implement the AI ​​cognitive level assessment method for any of the preceding brands.

[0074] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical applications, or technical improvements to the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for assessing a brand's AI awareness level, characterized in that, include: Obtain the industry and brand name of the brand to be evaluated; Based on the industry and the brand name, generate multiple keywords for the brand to be evaluated; For each of the aforementioned keywords, construct multiple Prompt questions; Multiple large AI models are used to answer each of the aforementioned Prompt questions, thus obtaining the answers to each of the aforementioned Prompt questions; Based on the answers provided, an assessment result is generated regarding the brand's cognitive level within the AI ​​ecosystem.

2. The method according to claim 1, characterized in that, When generating multiple keywords for the brand to be evaluated based on the industry and the brand name, the following are included: Obtain pre-built keyword prompt templates; The industry and brand name are injected into the keyword generation prompt template to obtain keyword generation prompts; Based on the keywords, prompt words are generated, and the AI ​​model is used to generate multiple keywords for the brand to be evaluated.

3. The method according to claim 1, characterized in that, When constructing multiple Prompt questions for each of the aforementioned keywords, including: Obtain a pre-built question prompt template; The industry and each of the keywords are injected into the question generation prompt template to obtain question generation prompts; Based on the question-generating prompts and the AI ​​big model, multiple prompt questions are constructed for each of the keywords.

4. The method according to claim 1, characterized in that, When generating the assessment result of the brand's cognitive level in the AI ​​ecosystem based on the answers provided, the following steps are included: The corresponding answer information is extracted from each of the answers, wherein the answer information includes at least one of the recommended brand name, search article information, and cited article information; Based on the answer information corresponding to each of the above answers, an assessment result of the brand's cognitive level in the AI ​​ecosystem is generated.

5. The method according to claim 4, characterized in that, When generating the assessment result of the brand's cognitive level in the AI ​​ecosystem based on the answer information corresponding to each of the aforementioned answers, the process includes: Based on the answer information corresponding to each of the answers, the index value of the preset cognitive level index is calculated, wherein the cognitive level index includes at least one of brand visibility, citation rate of brand promotion articles, and citation rate of the main text. Based on the index values ​​of the cognitive level indicators, an assessment result of the cognitive level of the brand to be evaluated in the AI ​​ecosystem is generated.

6. The method according to claim 1, characterized in that, After generating the assessment results of the brand's cognitive level in the AI ​​ecosystem, the following steps are also included: The results of the cognitive level assessment are then visualized.

7. The method according to claim 1, characterized in that, After generating the assessment results of the brand's cognitive level in the AI ​​ecosystem, the following steps are also included: Based on the assessment results of the brand's awareness level in the AI ​​ecosystem, the brand's promotional strategy in the AI ​​ecosystem will be adjusted.

8. A device for assessing a brand's AI cognitive level, characterized in that, include: The data acquisition module is used to obtain the industry and brand name of the brand to be evaluated; The keyword generation module is used to generate multiple keywords for the brand to be evaluated based on the industry and the brand name; The question generation module is used to generate multiple Prompt questions for each of the aforementioned keywords. The answer module is used to answer each of the Prompt questions using multiple large AI models, and obtain the answers to each of the Prompt questions; The evaluation module is used to generate an evaluation result of the brand's cognitive level in the AI ​​ecosystem based on each of the answers.

9. A branded AI cognitive level assessment device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to implement the method of any one of claims 1 to 7 when executing the executable instructions.

10. A non-volatile computer-readable storage medium storing computer program instructions thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images