Advertising method and device for equipment sales based on artificial intelligence

Abstract

The invention relates to the technical field of marketing, in particular to an advertising method and device for equipment sales based on artificial intelligence, and the method comprises the following steps: collecting social media behavior data and purchase records of consumers in a target market, and obtaining standardized consumer data through data cleaning and formatting processing; and performing user portrait analysis based on the standardized consumer data. According to the method and the device, the social media behavior data and the purchase record of the user are collected, and data cleaning and formatting processing are combined, so that the user portrait construction is more dynamic and accurate. Consumer subdivision not only depends on basic demographic information, but also introduces purchase frequency, interest labels and historical interaction behaviors to form more hierarchical user classification. Advertisement content adjustment is not limited to static text optimization, and visual elements, language styles and information presentation modes are dynamically adjusted according to sentiment analysis results, so that the content better fits preferences of different consumer groups.

Description

Technical Field

[0001] The present invention relates to the technical field of marketing, and in particular, to an advertising method and device for equipment sales based on artificial intelligence. Background Art

[0002] The technical field of marketing involves the promotion, sales, and distribution processes of products or services from producers to consumers, covering multiple aspects such as market research, consumer behavior analysis, brand management, marketing strategy formulation, advertising dissemination, and sales channel optimization. The advertising method for equipment sales is a promotion strategy in the technical field of marketing, aiming to improve the market awareness, customer interest, and sales conversion rate of equipment products through precise advertising content design and placement.

[0003] In the prior art, during the consumer segmentation process, user division is based on traditional demographic information, ignoring individual behavior characteristics and dynamic interest changes. This makes the advertising positioning vulnerable to the limitations of static data and difficult to accurately match user needs. The adjustment of advertising content mainly relies on fixed templates and fails to conduct personalized optimization in combination with user emotional feedback. The information expression method lacks pertinence, resulting in serious homogenization of advertising content and reducing user attention and click-through rate. Therefore, improvements are needed. Summary of the Invention

[0004] The purpose of the present invention is to solve the drawbacks existing in the prior art and propose an advertising method and device for equipment sales based on artificial intelligence.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions. The advertising method for equipment sales based on artificial intelligence includes the following steps: Collect social media behavior data and purchase records of consumers in the target market, and obtain standardized consumer data through data cleaning and formatting processing; based on the standardized consumer data, conduct user portrait analysis, including age, preferences, and purchasing power, to generate consumer segmentation results; Based on the consumer segmentation results, conduct sentiment analysis on the consumer group, identify the emotional reactions and preferences towards the advertising content, and obtain sentiment preference results; adjust the advertising copywriting and visual elements according to the sentiment preference results to generate customized advertising content; Use the customized advertising content to conduct tests on the platform, record user interaction data and feedback, and obtain advertising effect evaluation results; based on the advertising effect evaluation results, analyze the performance differences on different platforms, optimize the advertising placement strategy, and generate an optimized advertising deployment plan; According to the optimized advertising deployment plan, conduct advertising release, and at the same time monitor the real-time data stream, and adjust the advertising content and placement time according to the real-time data stream to obtain the placement results.

[0006] Preferably, the step of obtaining the standardized consumer data is as follows: Collect the social media behavior data and purchase records of consumers in the target market, remove duplicate data, and correct data format errors to obtain a preliminarily processed data set; Based on the preliminarily processed data set, identify and process missing values, outliers, and data biases, and apply data imputation and outlier correction to obtain the standardized consumer data.

[0007] Preferably, the step of obtaining the consumer segmentation result is as follows: Extract the age, purchase frequency, and preference categories of consumers from the standardized consumer data to obtain the feature extraction result; Based on the feature extraction result, calculate the purchasing power index of each consumer. The calculation formula is: Wherein, represents the purchasing power index of consumer , represents the age value of consumer , represents the median of the age values of all consumers, represents the purchase frequency of consumer , represents the preference category of consumer , represents the total sum of the purchase frequency values of all consumers in the target market, represents the total purchase amount of consumer ; Based on the purchasing power index, combine the purchase frequency, preference category, and age range of consumers to judge the category attribution of consumers and generate the consumer segmentation result.

[0008] Preferably, the step of obtaining the emotional preference result is as follows: Extract the advertising interaction data of each group from the consumer segmentation result, including the number of likes, the number of comments, the sharing frequency, and the advertising viewing duration, and perform data screening and classification to obtain the interaction characteristics of consumer groups; Based on the interaction characteristics of consumer groups, calculate the emotional response score of each group to the advertising content. The calculation formula is: Wherein, represents the emotional response score of group , represents the number of likes of group , represents the number of comments of group ; Advertising sharing frequency of the representative group and the advertising viewing duration of the representative group; Advertising viewing duration of the representative group Based on the emotional response score and combined with the advertising interaction patterns of consumer groups, analyze the emotional trends of each consumer group towards the advertising content to obtain the emotional preference results. Preferably, the steps for obtaining the customized advertising content are as follows:

[0009] According to the emotional preference results, extract the tendencies of different consumer groups towards the advertising text and visual design to generate advertising copywriting features and visual element features; Based on the advertising copywriting features and visual element features, screen the advertising text content and visual materials that meet the emotional preferences of the consumer groups, and adjust the expression of text sentences, the presentation method of advertising themes, and the information transmission structure to generate an advertising copywriting adjustment plan and a visual element adjustment plan; Based on the advertising copywriting adjustment plan and the visual element adjustment plan, integrate the optimized copywriting content and visual materials, and adjust the matching degree between the advertising content and the characteristics of the consumer groups to generate customized advertising content.

[0010] Preferably, the steps for obtaining the advertising effect evaluation results are as follows: Use the customized advertising content to conduct an advertising placement test on the platform, set the advertising placement strategy, including the release time, display frequency, and placement audience range, collect the user interaction behavior data, and generate a user interaction data set; Based on the user interaction data set, calculate the advertising adaptability score, and the calculation formula is: wherein, represents the advertising adaptability score, represents the advertising click-through count, represents the advertising residence time, represents the advertising sharing count, represents the advertising comment count, represents the advertising display count; Based on the advertising adaptability score, analyze the performance of the advertising on the platform, compare the user interaction effects of each advertisement, and obtain the advertising effect evaluation results.

[0011] Preferably, the steps for obtaining the optimized advertising deployment plan are as follows: Based on the advertising effect evaluation results, calculate the performance differences of advertising placements on each platform, analyze the user behavior patterns of different platforms, compare the effects of advertising content among different audience groups, and generate an optimized advertising placement strategy plan; Based on the optimized advertising placement strategy plan, adjust the advertising placement time, adjust the display methods of advertising content on each platform, optimize the adaptation strategy of advertising materials, and adjust the budget allocation ratio to form an optimized advertising deployment plan.

[0012] Preferably, the steps for obtaining the placement results are as follows: According to the optimized advertising deployment plan, conduct advertising placements, monitor the user behavior data after the advertisement is exposed, and generate real-time advertising feedback data; Based on the real-time advertising feedback data, calculate the advertising placement results, and the expression is: where represents the advertising placement result of advertisement , represents the user page interaction intensity of advertisement within time , represents the content browsing depth of advertisement , represents the page bounce rate of advertisement , represents the average stay duration of advertisement in each time period, represents the number of repeated visits of users to advertisement , and T represents the total duration after the advertisement is placed; Based on the advertising placement results, analyze the user response of the advertisement on each platform and time period to obtain the placement results.

[0013] The present invention provides an advertising device, including: a processor and a memory. The memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory, so that the advertising device executes the above-mentioned advertising method for equipment sales based on artificial intelligence.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In the present invention, by collecting users' social media behavior data and purchase records, and combining data cleaning and formatting processes, the construction of user portraits becomes more dynamic and accurate. Consumer segmentation not only relies on basic demographic information but also introduces purchase frequency, interest tags, and historical interaction behaviors to form a more hierarchical user classification. The adjustment of advertising content is not limited to static text optimization. Instead, according to the results of sentiment analysis, visual elements, language styles, and information presentation methods are dynamically adjusted to make the content more in line with the preferences of different consumer groups. In the advertising placement stage, a platform testing mechanism is adopted to record users' interaction data and, combined with feedback information, establish a multi-dimensional advertising effect evaluation system to enable the placement strategy to have dynamic adaptability. For advertising deployment, a real-time data stream monitoring mechanism enables the placement plan to have self-adaptive adjustment capabilities. According to market feedback, the display method, release time, and interaction guidance strategy are adjusted to improve the response efficiency of advertising content. The whole-process data closed-loop makes the advertising dissemination path more visual and optimizable, reducing resource waste and increasing advertising coverage and conversion rates. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of the steps of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0016] In order to make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0017] Please refer to Figure 1 , the present invention provides a technical solution, an advertising method for equipment sales based on artificial intelligence, including the following steps: Collect the social media behavior data and purchase records of consumers in the target market, and through data cleaning and formatting processes, obtain standardized consumer data; based on the standardized consumer data, conduct user portrait analysis, including age, preferences, and purchasing power, to generate consumer segmentation results; Based on the consumer segmentation results, conduct sentiment analysis on the consumer groups to identify the emotional reactions and preferences towards the advertising content, and obtain sentiment preference results; according to the sentiment preference results, adjust the advertising copy and visual elements to generate customized advertising content; Use the customized advertising content to conduct tests on the platform, record the interaction data and feedback of users, and obtain advertising effect evaluation results; based on the advertising effect evaluation results, analyze the performance differences on different platforms, optimize the advertising placement strategy, and generate an optimized advertising deployment plan; According to the optimized advertising deployment plan, conduct advertising releases, and at the same time monitor the real-time data stream, and adjust the advertising content and release time according to the real-time data stream to obtain the placement results.

[0018] The steps to obtain standardized consumer data are: Collect social media behavior data and purchase records of consumers in the target market, remove duplicate data, correct data format errors, and obtain a preliminarily processed data set; Based on the initially processed data set, missing values, outliers and data deviations are identified and processed, and data interpolation and outlier correction are applied to obtain standardized consumer data.

[0019] Specifically, based on the previously acquired consumer social media behavior data and purchase records, we first check whether there are consecutive records of the same user according to the pre-set duplication judgment criteria. For example, when the user ID is the same and the record time difference is within seconds, it is considered repeated. The second threshold is derived from the average value of repeated samples collected in the past plus The experience redundancy of seconds is used to traverse and compare all data entries in this way, and the records that meet the above duplication conditions are eliminated. Then, the format of the text field and the numeric field is checked to see if it meets the agreed specifications, such as whether the date format meets The format of the date or the value field contains non-numeric characters. If confusion is found (for example, the month and date are swapped or the time zone information is wrong), they are compared one by one according to the pre-prepared time calibration comparison table, in which common date and time zone combinations and their matching ratios are recorded. When the matching ratio is greater than If the format is correctable, the string will be split according to the manually set segmentation rules for comparison. If a symbol error is detected in the numeric field (such as an extra "," or "."), it will be directly removed or replaced with an integer. The judgment threshold of the integer form here is derived from the statistical distribution of common currency amounts in the target market. Assume that The price value falls within to If a value exceeds this range, it is considered as a suspected format abnormality. Then, combined with the user's purchase volume or transaction type, it is determined whether it needs to be corrected or eliminated. Finally, all records that have undergone repeated checks and format corrections are collected together to obtain a preliminary processed data set.

[0020] Based on the preliminarily processed data set obtained above, we first scan each item to identify whether there are missing values. When it is found that a field is empty or the symbol is occupied and the content cannot be confirmed, we interpolate it with reference to the historical average value with the same meaning as the field. The historical average value is a comprehensive calculation of the median and mean selected from the same type of field data in the past three months. For example, for a single amount field, the mean can be calculated in the range of 0 yuan to 10,000 yuan and the compensation value can be calculated in combination with the transaction frequency so that the interpolated value has a reasonable distribution. Then we continue to detect outliers and data deviations, and compare each record with the preset upper and lower limits. For example, if the age is limited to to If the range is exceeded, the quartile method is used to check the difference from the normal distribution. If the range exceeds a certain multiple of the sum of the distance between the third quartile and the first quartile (the multiple is evaluated by the common extreme value positions in the training set, and the empirical setting is ) is temporarily identified as an abnormal record, and then combined with the number of purchases or preference types of the user to determine whether to keep it. If it is determined to be abnormal, it can be further corrected by weighted interpolation, such as making corrections based on the mean and variance of the same group of people on this field. The interpolation weight threshold will be calculated during correction. for ,Should The value comes from the prediction deviation test results of the model on 10,000 samples. If deviation still occurs after interpolation, the record will be marked as pending verification. All entries that have completed missing and abnormal processing are summarized after this process is completed, and the data interpolation outlier correction operation is completed immediately to obtain standardized consumer data.

[0021] The steps to obtain consumer segmentation results are: Extract consumers’ age, purchase frequency, and preferred categories from standardized consumer data to obtain feature extraction results; Based on the feature extraction results, the purchasing power index of each consumer is calculated using the following formula: in, Representing consumers Purchasing power index, Representing consumers The age value, Represents the median age of all consumers. Representing consumers The purchase frequency, Representing consumers Types of preferences, Represents the sum of the purchase frequency values ​​of all consumers in the target market. Representing consumers Total purchase amount; Based on the purchasing power index, combined with consumers' purchase frequency, preferred categories and age range, consumers are classified and classified to generate consumer segmentation results.

[0022] Specifically, based on the age, purchase frequency and preferred category information marked in the standardized consumer data obtained earlier, each record is concatenated according to the consumer's unique number, and the corresponding age field, purchase frequency value and preferred category mark are read. In this process, a classification comparison table is first referred to to determine the attribution of the preferred category. The comparison table is mapped by common product types and attributes in the target market. When the product label of a record matches a category in the comparison table by more than 80%, the consumer's transaction is classified into this category. When the matching degree is less than 80%, it is further compared according to the lower-level classification directory. After confirming the preferred category, the transaction is summarized. The total number of purchases is calculated and the age field is recorded. The age field is compared with the preset valid range of 0 to 120 years old. When a consumer's age is less than 0 or greater than 120, the record comparison operation is performed. The adjacent records under the same number and the global statistical data are searched to determine whether it is an input deviation. If the input deviation is confirmed, the record is eliminated. The purchase frequency value is also limited to between 1 and 1000 times and checked. Exceeding this range means that the corresponding record has an extreme abnormality. The entire transaction data under the number can be queried again based on its frequency of occurrence for verification. Finally, all records that meet the conditions are packaged and summarized to obtain the feature extraction results.

[0023] The benefit of the formula is that it combines the age difference item and the purchasing behavior item in the numerator and introduces the total purchase frequency and the square root of the consumption amount in the denominator to comprehensively measure the performance of consumers in different dimensions. This allows consumers whose ages deviate more from the median age of market entities and whose purchasing behavior is more active to present higher or lower values ​​in the purchasing power index. This numerical difference helps to accurately distinguish different types of consumer groups in subsequent segmentation links.

[0024] The steps to obtain the parameters are as follows: The age value is directly read from the standardized consumer data obtained above. The age value is obtained through ID card registration information or information actively filled in by the user, and is deduplicated and processed when collected. For some parallel inputs (for example, when the same consumer fills in conflicting ages on different occasions), the actual registered valid ID records are retrieved for comparison. If the age field is found to be outside the range of 0 to 120 years old, it will be regarded as an input anomaly, and then corrected or eliminated by comparing with adjacent records. In a test of 500,000 registered users, about 2% of the age field values ​​of the population deviated from the reasonable range, and finally a stable range was obtained by using the corresponding correction method. value

[0025] The steps to obtain the parameter are as follows. It is the median of all consumers' age values, and it is necessary to sort the complete list of user ages and take the median value. Suppose a total of N valid age records are collected. When N is odd, take the age value at the middle position after sorting. When N is even, take the average of the values at the two middle positions. For example, in the scenario of collecting age data of 600,000 consumers, the ages at the 300,000th and 300,001st positions after sorting all ages are 32 and 33 respectively. Then = = 32.5

[0026] The steps to obtain the parameter are as follows. It represents the consumer purchase frequency within a specified period, which can be obtained by querying the historical transaction database. Summarize all the orders of this consumer within this period. If a consumer has completed 45 orders within a year, then = 45. When further analyzing, the order timestamp can be combined. Multiple orders on the same day will be counted independently for each order. By counting the order data of 500,000 users within a year, accumulate the number of transaction times of each user in this interval, and then form a purchase frequency array, where the purchase frequency value of each user is used as the specific value of this parameter

[0027] The steps to obtain the parameter are as follows. It is the type of preference of the consumer First, classify the types of goods involved by the consumer in multiple orders, and then map the types of goods to numerical values. For example, assign the value 2 to the electronic product category, assign the value 3 to the home life category, assign the value 5 to the catering and food category, etc. If a certain consumer purchases the same category of goods multiple times within a specified period, then take the category with the highest frequency as the main preference of this consumer at the current stage, and use this numerical value as , for example, a certain consumer has placed 50 orders within a year, among which 30 orders are concentrated in electronic products, 12 orders are for catering and food, and 8 orders are for other categories. Then their = 2

[0028] The steps to obtain the parameter are as follows. It represents the total sum of the purchase frequency values of all consumers in the target market. First, select the analysis period, and then summarize the purchase frequency of each consumer within this period one by one. After the summarization is completed, add them up. For example, when conducting statistics on 300,000 consumers within a one-year period, the total number of order pens of all consumers adds up to 6,000,000 pens. Then = 6,000,000

[0029] The steps for obtaining the parameter are as follows. It represents the total purchase amount of consumers. First, retrieve the payment amounts of all orders of a consumer within a specified period from the transaction database, and then accumulate these payment amounts item by item to obtain a total value. For example, in the order data of a consumer, the total payment amount of 30 orders is 3,000 yuan, the total payment amount of 15 orders is 1,800 yuan, and the total payment amount of the remaining 5 orders is 450 yuan. Then, for this consumer within the specified period = 3,000 + 1,800 + 450 = 5,250. =3000+1800+450=5250。

[0030] Calculation process: Let the relevant parameter values of a certain consumer be , , , , , , first calculate = = 7, and then = . Next, find = 30 + 2 = 32. The numerator part is . The denominator part first calculates = . Add it to to get = 500,059.1608. Thus, we get: This result shows that in this example, the purchasing power index value of this consumer is approximately 0.0001693. When the same calculation is performed for all consumers in the same market dimension, different intervals can be set for the obtained sequence of purchasing power indices for segmented comparison. When the value of a certain consumer is significantly higher than the overall average, it indicates that there is a significant difference between their age and the median, and their purchasing behavior is relatively active. A too low value means that this consumer is not prominent in terms of activity or amount contribution. Therefore, in the subsequent steps, consumers can be classified into different categories based on these values.

[0031] Based on the previously obtained purchasing power index, combined with the purchase frequency of consumers and the corresponding preference categories, as well as the age range labels established by the age field, a classification criterion needs to be established before starting the category attribution judgment. This criterion usually sets several intervals, and each interval contains a range of purchasing power indices and one or more preference category values. First, locate the consumer's purchasing power index within the known group of intervals. For example, divide the purchasing power index into several segments with a step of 0.0001. When a consumer's value is between 0.0001 and 0.0003, it can be regarded as the middle paragraph. When the purchase frequency is between 20 and 40 and the preference category is equal to 2 or 3, it is judged as one group. When the purchasing power index is close to 0.001 or higher, it is classified into a higher activity group, and when the purchasing power index is lower than 0.00005, it is included in another group. Combining the distinction of age ranges, multiple different population dimensions can be distinguished. After the category attribution judgment of consumers, the subdivision labels of each consumer are generated. Once it is found that the parameters are updated or new changes occur in the consumer behavior data, the classification process can be re-executed and the change trajectory of the consumer's attributed category in different time periods can be compared to finally obtain the consumer segmentation result.

[0032] The steps to obtain the emotional preference result are as follows: Extract the advertising interaction data of each group from the consumer segmentation result, including the number of likes, the number of comments, the sharing frequency, and the advertising viewing duration, perform data screening and classification to obtain the interaction characteristics of the consumer group; Based on the interaction characteristics of the consumer group, calculate the emotional reaction score of each group to the advertising content. The calculation formula is: Among them, represents the emotional reaction score of group , represents the number of likes of group , represents the number of comments of group , represents the advertising sharing frequency of group , represents the advertising viewing duration of group ; Based on the emotional reaction score, combined with the advertising interaction mode of the consumer group, analyze the emotional trend of each consumer group towards the advertising content to obtain the emotional preference result.

[0033] Specifically, based on the group information indicated in the previously obtained consumer segmentation results, cross-reference each group record one by one, read the like count field from it, and determine whether it is too high or too low according to a preset numerical range. For example, compare the like count with the range of 0 to 500 times, and mark the group with a like count significantly exceeding this range as an outlier. Establish the same interval comparison method for the comment count, compare the comment count with the range of 0 to 1000 comments, and find that any group with more than 1000 comments is regarded as having an abnormal concentration of comments. Then, retrieve the advertising sharing frequency of this group and compare it with the threshold range between 1 and 50 times. When there is a value higher than 50 times, trace the actual scenario according to the sharing records of specific members within the group. The advertisement viewing duration is compared with the interval of 3 to 600 seconds. During the inspection process, once missing data is found in the like count, comment count, sharing frequency, or viewing duration, it is marked in a timely manner. Subsequently, classify all data entries within the normal range. For example, classify the groups with like counts and comment counts in the middle range into one category, and separately classify the groups with significantly higher sharing frequencies but smaller viewing durations into another category. After classification, separate and manage the outlier groups from other groups. Finally, summarize the performance combinations of each group in terms of likes, comments, sharing, and viewing duration to obtain the interactive characteristics of consumer groups.

[0034] The advantage of the formula is that by adding the square of the like count and the comment count to the numerator part, introducing the square root of the advertising sharing frequency and adding a reference value in the denominator, and combining the logarithmic operation of the viewing duration, it can comprehensively reflect in a single score whether the like behavior is relatively concentrated, whether the comment interaction is active, the intensity of sharing activities, and the viewing duration distribution, enabling the core interactive characteristics of different groups to be quantitatively measured within the same numerical framework.

[0035] The steps to obtain the parameter are as follows. It represents the group The total number of likes within the specified delivery period. First, read the like behavior occurrence entries from all individual user records within the group, and accumulate the number of likes one by one in the complete interaction data using the user's unique identification information as an index. The recording objects include various dimensions such as text content likes, image likes, and short video likes. After the statistics are completed, add up these like cumulative values to form , in a monitoring of 20,000 advertising interaction samples, update the like entries on a daily basis, and confirm the total number of likes that each group has in the 7-day delivery period. For example, in the statistical result of the user set with group number 3, it shows that this group has generated 260 likes within 7 days, then = 260. = 260.

[0036] The steps to obtain the parameter are as follows. It represents the group The number of comments received needs to be retrieved from the comment database and merged based on the comment behaviors of all users in the group. First, lock the group The identifiers corresponding to the contained users and search for the comment posting records of the same identifiers in the specified advertising environment. Sum up the number of comments in all relevant records to obtain , for example, for users with group number 3, the set monitoring range is one month. If the calculated value of the sum of comment entries within one month reaches 120, then = 120.

[0037] The steps to obtain the parameter are as follows. It represents the group The sharing frequency during the advertisement delivery period. When obtaining it, first count the number of times all users in the group share the target advertisement, then divide the sharing times by the monitoring days to obtain a daily average sharing value, and multiply the daily average sharing value by the total number of days of cumulative exposure of this advertisement in this group to form the sharing frequency , in a 10-day trial delivery, if a total of 90 shares occurred under group number 3, then the daily average sharing times is 9. Since the advertisement delivery days are 10 days, so = 9 × 10 = 90.

[0038] The steps to obtain the parameter are as follows. It represents the group The viewing duration of the advertisement content. When recording, calculate the time point when each user opens the advertisement content and the time point when they close the advertisement content to obtain the number of seconds for a single viewing, and then accumulate the viewing seconds of all users in this group within a given period to obtain , for example, in the record of a 14-day advertisement delivery, the total viewing duration of all users with group number 3 is 5400 seconds, then = 5400.

[0039] Calculation process: Let , , , , first calculate the numerator part = = 900 + 12 = 912, and then the denominator part = = , to obtain: Then calculate = = , add the two together, that is: This result shows that in the current sample, the group The emotional response score of different groups is about 349.9563. When comparing the value with this value, the group with a higher value tends to be more active in likes, comments, sharing and viewing time, while a lower value suggests that the group has a lower degree of investment in these interactive dimensions. Adopt corresponding promotion strategies or content optimization ideas in different intervals.

[0040] After obtaining the emotional response scores of each group, a specific comparison is carried out on each group based on the pre-organized advertising interaction pattern classification list. First, the strength level of the click behavior and stay behavior that each group has shown is read, and then the distribution of likes, comments and shares made by the group in the past period of time is used for reference. When it is found that the emotional response scores of some groups are between 300 and 400 and the number of comments has continued to grow in the near future, this group will be classified as an emotional trend with a focus on positive evaluation. If the emotional response scores of some groups are between 100 and 200 and the corresponding number of likes increases slightly but the number of comments remains in a stable range , it is regarded as a neutral emotional trend. In addition, when the emotional response score is lower than 50, it will be confirmed in combination with the viewing time and sharing frequency characteristics of the group. Once the viewing time and sharing frequency are at a low level between 5 and 10 times, it will be classified as a low-investment group. At the same time, for emotional response scores above 600, by comparing their overall interaction indicators, such as more than 800 likes or more than 500 comments, it shows that the group shows a stronger advertising relevance. After completing the cross-statistics of all emotional response scores and interaction patterns, the emotional type and tendency of each group towards the advertising content during this stage are recorded one by one to obtain the emotional preference results.

[0041] The steps to obtain customized advertising content are: Based on the results of emotional preference, extract the inclinations of different consumer groups towards advertising text and visual design, and generate advertising copy features and visual element features; Based on the characteristics of advertising copy and visual elements, screen the advertising text content and visual materials that meet the emotional preferences of consumer groups, adjust the expression of text sentences, the presentation of advertising themes and the information transmission structure, and generate advertising copy adjustment plans and visual element adjustment plans; Based on the advertising copy adjustment plan and the visual element adjustment plan, the optimized copy content and visual materials are integrated, the matching degree between the advertising content and the characteristics of the consumer group is adjusted, and customized advertising content is generated.

[0042] Specifically, according to the emotional preference results obtained previously, after reading the corresponding characteristics of different consumer groups marked therein, first compare the reference data of the group in terms of reading habits and color acceptance. Check one by one through the previously established association comparison table whether there are resistance values to specific modifiers or preference intensity records for certain visual elements in the group. If it is found that the resistance value is greater than 3, it means that the corresponding vocabulary will cause obvious rejection emotions within the group. The threshold of 3 is obtained by investigating the negative reaction rates when 1000 groups use words with different emotional colors and conducting statistics. At the same time, screen the entries within each group with a color acceptance greater than 60% for bright colors and rearrange them. The empirical threshold of 60% is obtained by one-to-one mapping of past advertising data. When the mapping results show that the reach of a certain color by the majority of people exceeds 60%, it is regarded as having an obvious acceptance tendency. Then, classify this analysis information. For example, group the groups with a more straightforward expression style and stronger main colors into one tendency group, and classify the groups with a more gentle narrative logic and more neutral color matching into another tendency group. After grouping, further cross-group comparison of the extracted modifier words, descriptive phrases, color matching schemes, etc. can be carried out to confirm the overlapping parts. Once it is found that certain words or visual design forms are significantly attractive in multiple groups, mark them as general elements and record them in the database. When all important elements such as words, colors, and background graphics are marked, they can be matched with the specific tendencies of each group. If a group shows a greater acceptance of warm-toned colors and a stable positive feedback to promotional copywriting, write the preferred combination of such copywriting and visual composition in the mark. Finally, integrate the tendencies of all groups to form the corresponding advertising text selection criteria and visual configuration instructions, and generate advertising copywriting characteristics and visual element characteristics.

[0043] Based on the characteristics of advertising copywriting and visual elements, first select several text patterns that have a high degree of coincidence with the emotional preferences of the consumer group and establish a pattern index table internally. The entries in this index table include specific modifiers, paragraph order, and potential illustration style markers. When encountering expressions with overly inflammatory words or conflicting with the previous threshold judgment, immediately determine whether to eliminate them according to the previously calculated emotional rejection level. If the rejection level exceeds 1, it indicates the existence of inappropriate elements and needs to be removed or replaced in the pattern. This rejection level is the ratio of the number of negative reviews to the total number of readings after collecting more than 200 negative reviews. Through this comparison, an average proportion is obtained, and a proportion exceeding 1 is regarded as not suitable for use. After finishing the arrangement, the presentation methods of the advertising theme also need to be compared one by one. For example, for the same theme, use methods such as mixed text and graphics or preview of short video clips, and compare the reading duration, click-through rate, and interaction rate of each consumer group in parallel. When it is found that a certain method can maintain a high click rate among most groups but the number of comments is consistently less than 3 for a long time, mark it as a presentation method with strong attractiveness but insufficient interaction depth. In the information transmission structure part, it is necessary to detect whether there are long text paragraphs that cause users to exit during reading, and count and compare whether the proportion of samples with a page stay time greater than 10 seconds remains above 70%. If it is lower than 70%, it means that the information level needs to be reduced or the order needs to be adjusted. When adjusting, place the key text first and use colors or graphics to distinguish it to prevent visual congestion. Finally, match the text content verified by the comparison table with the picture materials respectively, and list the text-visual comparison positions for each matching scenario to generate an advertising copywriting adjustment plan and a visual element adjustment plan.

[0044] Based on the advertising copywriting adjustment plan and the visual element adjustment plan, first read the appropriate patterns from the text pattern index table and perform corresponding splicing with the marked visual materials. Package the color presets that are known to be consistent with the group characteristics under the same theme framework, and at the same time assign a certain order to the key text paragraphs in the information transmission structure so that users can see the core content as soon as they enter the page. When integrating, it is necessary to check whether the contrast between the color scheme and the text meets the previously set readability requirements. This readability requirement is derived from measuring the text reading accuracy of 150 users and regarding the color scheme with a contrast ratio greater than 4.5:1 as a better plan. If it is found that there is a background conflict in the overlapping area of the existing illustration and the text, reselect other materials or add color blocks at the bottom of the image. After all parts are matched, write the information such as the text presentation position, image proportion, and theme matching into the configuration list item by item, and use visual inspection to determine the display size and layout distribution ratio. Also record item by item the possible differences in font size or layout in a multi-terminal environment. When the final text combination and visual materials meet the tendencies of each group and there are no reading obstacles or content ambiguities in the trial preview session, confirm that this group of plans can be used for multi-platform delivery and generate customized advertising content.

[0045] The steps to obtain the advertising effect evaluation results are as follows: Use customized advertising content to conduct advertising placement tests on the platform. Set advertising placement strategies, including release time, display frequency, and target audience scope. Collect users' interaction behavior data to generate a user interaction dataset; Based on the user interaction dataset, calculate the advertising fitness score. The calculation formula is: where, represents the advertising fitness score, represents the number of clicks on the advertisement , represents the dwell time of the advertisement , represents the number of shares of the advertisement , represents the number of comments on the advertisement , represents the number of displays of the advertisement ; Based on the advertising fitness score, analyze the performance of the advertisement on the platform, compare the user interaction effects of each advertisement, and obtain the advertising effect evaluation results.

[0046] Specifically, according to the customized advertising content obtained previously, before performing the placement test, establish a basic list of placement strategies. Set the release time within a twelve-hour range from 9:00 to 21:00 every day and segment it at 30-minute intervals. Limit the display frequency to a maximum of 3 times within each time period and record the page dwell situation after each exposure. Target the advertising audience at registered users aged 18 to 45 years old and further screen them according to interest tags. When placing different advertising contents, check the overlap between the release time period and the corresponding population item by item to avoid repeated exposures exceeding 5 times. After starting the placement, monitor the click count, comment entries, and possible sharing behaviors in real time. Statistically compare the placement progress within each time period. If it is found that the click count is greater than 100 times or the number of comment entries exceeds 10, mark this time period as a high-activity section. These empirical thresholds of 100 times and 10 are obtained from the aggregated average value of 10,000 previous placement data and are corrected in combination with the results of one on-site observation. During the statistical process, summarize all relevant user browsing time information and associate it with their interaction actions. Record the interaction behavior data line by line to form a user interaction dataset. After the placement cycle ends, uniformly check the viewing completion rate of users for the advertisement and the page bounce distribution situation, supplement complete interaction index items, and finally merge all behavior statistics items to form the final data and generate a user interaction dataset.

[0047] The advantage of the formula is that it combines the number of clicks and the dwell time in the numerator part, and by taking the logarithm of the number of shares, the extreme share values are prevented from being overly amplified. At the same time, the ratio of the number of comments to the number of displays is independently added, enabling the balanced measurement of interaction behaviors in different dimensions within the same scoring system.

[0048] The steps to obtain the parameter are as follows. It represents the advertisement The number of clicks during the placement period. First, record the total number of times the advertisement is clicked by users in each time period, and then add up the click volumes in each time period at the end of the day to form the total number of clicks for the day. Accumulate this to the end of the placement cycle to obtain , for example, if the daily click volumes of a certain advertisement during a 7-day placement cycle are 50, 60, 55, 40, 75, 65, and 80 times respectively, then = 50 + 60 + 55 + 40 + 75 + 65 + 80 = 425.

[0049] The steps to obtain the parameter are as follows. It represents the advertisement The dwell time. It is obtained by collecting the difference between the start browsing time and the end browsing time of users on the advertisement page and accumulating it. At the same time, a low-limit reference value of 3 seconds is set. When the dwell time is less than 3 seconds, it is still regarded as a single browsing but classified as "extremely short dwell". This 3-second reference value is obtained by taking the median of the average reading start time of 1000 users on the graphic page. When the accumulated dwell time of this advertisement reaches 30000 seconds within 7 days, then = 30000.

[0050] The steps to obtain the parameter are as follows. It represents the advertisement The number of shares. First, read the sharing behavior records of all users during the advertisement display period and lock the unique identifier of the advertisement. If the advertisement is shared a total of 200 times during the 7-day placement, then = 200.

[0051] The steps to obtain the parameter are as follows. It represents the advertisement The number of comments. It is obtained by summarizing the number of comment data entries submitted by users under the advertisement page. Add up the daily number of comments. It is also necessary to perform duplicate identification on the comment content. If the same user posts the same comment within a very short interval, it is marked as one. For example, if the cumulative number of comments of a certain advertisement during this cycle is 90, then = 90.

[0052] The steps to obtain the parameter are as follows. It represents the advertisement The number of impressions, by recording the exposure frequency of the advertisement at each time period, considering multiple refreshes by the same user in the same time period as one impression. If the advertisement is displayed 2000 times by the system after the statistical period ends, then = 2000.

[0053] Calculation process: Let , , , , , first calculate = , add it to to get , then take the natural logarithm of = 200 + 1 = 201, obtaining the first term's numerator divided by the denominator: Then calculate the second term = = = , after adding the two terms ; This result indicates that when is above 100, this advertisement is relatively active in terms of click performance, dwell time, sharing, and comments. If the values of all advertisements are aggregated and sorted, the interaction intensities of different advertisements can be compared. The higher the value, the more attention it receives at the user level, and too low a value means its attractiveness or interactivity is limited. These quantitative data can help complete subsequent differential analysis and optimization deployment.

[0054] After calculating the advertising adaptation score, first collect the score lists of each advertisement and place them side by side with the corresponding placement parameters. Read the key interaction indicators of each advertisement during the placement period, such as click-through rate, dwell time, and comment trend, etc. Compare with other advertisements placed during the same time period to check the gap between the high-activity section and the low-activity section. Re-check the advertisement entries with scores exceeding 200 from dimensions such as the number of clicks exceeding 500 times and the number of comments exceeding 100 to confirm whether the advertisement shows a significant increase in both user attention and subsequent sharing willingness. Mark the advertisements with scores below 50 as those with insufficient attention and check whether the ratio of display times to clicks is less than 1:100. If the ratio is less than this value, it means that the display volume of the advertisement is much larger than the interaction response volume, and the comment and sharing records can be further traced. After confirming that there are no duplicate statistics or abnormal data in the specific score sequence, number all the advertisements from high to low scores and conduct data visualization inspection item by item. Finally, compile an overall comparison chart of the platform's advertising interaction level, list all scores and core indicators and mark the difference intervals to obtain the advertising effect evaluation result.

[0055] The steps to obtain the optimized advertising deployment plan are as follows: Based on the advertising effect evaluation result, calculate the performance differences of advertising placements on each platform, analyze the user behavior patterns of different platforms, compare the effects of advertisement content among different audience groups, and generate an optimized advertising placement strategy plan; Based on the optimized advertising placement strategy plan, adjust the advertising placement time, adjust the display methods of advertisement content on each platform, optimize the adaptation strategy of advertisement materials, and adjust the budget allocation ratio to form an optimized advertising deployment plan.

[0056] Specifically, based on the advertising effect evaluation results, first classify all the obtained interaction metrics by platform. Split the exposure duration, click-through times, and comment counts of the same advertisement on different platforms according to the audience type. Read the user behavior records of each platform and compare each record with the pre-established valid range. For example, compare the daily active user volume with the range of 1,000 to 50,000, compare the average stay time with the range of 10 seconds to 300 seconds, and compare the interaction frequency with the range of 0 to 10 times. If it is found that the record exceeds the range, mark the record as an outlier and conduct a data check. After confirming that there is no error, summarize and form a platform difference comparison item. Further disassemble the audiences of each platform in combination with the comment preference degree and sharing tendency. For example, in some social platforms, when the number of comments is greater than 100 but the number of likes is less than 20, it is determined as a mode with active comments but low click-through conversion. In another part of the information platforms, a consistent interaction phenomenon exceeding the threshold is found. The threshold is set by counting the extreme values and calculating the interquartile range in 10,000 advertisement records. When the data distribution is outside Q1 and Q3 by more than 1.5 times the range of Q3 - Q1, it is regarded as extreme. Finally, conduct a parallel measurement of the click-through rate, comment rate, and stay duration between platforms. If some platforms show significantly higher clicks on specific copywriting or color matching, and the interaction peak of some platforms during night-time placement is greater than that during the day, it can be judged that the audience states of these platforms at different times vary significantly. After all the comparisons are completed, refer to the corresponding population portraits and conduct a cross-analysis of the response degrees of the same age group or the same interest tags. Map the performance results of each platform to the original advertising effect evaluation results, integrate the weighted ratio of clicks and comments, and rank the platforms based on this. After obtaining the ranking, further compare the user behavior characteristics of the high-ranking and low-ranking platforms, record these difference data, and then summarize to obtain the differences in the placement performance of each platform. Combine the feedback of different audience groups on different platforms, determine the priority order of the platforms according to the click and feedback intervals obtained earlier, make specific adaptation instructions for the placement duration and material styles, and generate an optimized advertising placement strategy plan.

[0057] Based on the advertising placement strategy optimization plan, first adjust the advertising placement time for each platform and re-divide the display segments. Define the period from 6 am to 9 am as segment A, from 9 am to 6 pm as segment B, and from 6 pm to 12 am as segment C. According to the previous statistical results of the active peaks of each platform, determine which segments are most suitable for placing main product type advertisements and which segments are more suitable for displaying emotional appeal content. After confirming each segment, associate the copywriting and visual materials of the advertising content. For example, allocate long narrative videos to social platforms that prefer videos, and focus on key titles and match with a small number of pictures on information platforms that prefer short text reading. When setting the budget allocation ratio, first retrieve the difference between the previous placement cost and the interaction revenue. If the difference is between 50% and 80%, it indicates that the investment can be further increased. This range is referenced by the statistical median value of advertising expenditure and recovered revenue. When the actual placement rate exceeds 80% of the median value, it is recommended to lower the placement budget to control the marginal investment. For the case where the difference is less than 20%, it means the investment is relatively conservative and the budget can be appropriately increased. After corresponding adjustments have been made to the budget allocation and display methods for all platforms, the color scheme of each advertisement within the corresponding time period should also be cross-reviewed to ensure that the color contrast meets the readability standards collected previously. Before the advertisement goes live, re-arrange the display frequency and placement content for each platform, and at the same time limit the number of words in the advertisement title, for example, control it within 15 to 30 words. Write all the above configurations into a centralized document, and finally merge the placement time arrangement and the corresponding method of advertising materials to form an optimized advertising deployment plan.

[0058] The steps to obtain the placement results are as follows: According to the optimized advertising deployment plan, conduct advertising placement, monitor the user behavior data after the advertisement is exposed, and generate real-time advertising feedback data; Based on the real-time advertising feedback data, calculate the advertising placement result, and the expression is: Among them, represents the advertising placement result of the advertisement, represents the user page interaction intensity of the advertisement within the time , represents the content browsing depth of the advertisement , represents the page bounce rate of the advertisement , represents the average stay duration of the advertisement in each time period, represents the number of repeated visits of the user to the advertisement , and T represents the total duration after the advertisement is placed; Based on the advertising delivery results, analyze the user response of the advertisement on each platform and time period to obtain the delivery results.

[0059] Specifically, according to the optimized advertisement deployment plan obtained previously, at the beginning of the delivery, sort out the exposure positions and display time periods of each advertisement, match the delivery time of the same advertisement on different platforms with the population information, and record the page opening situation and interaction click flag under each exposure at the same time. If the exposure volume within a certain delivery duration is greater than 100 times and the click quantity is less than 3 times, it is marked as a low conversion period. The values of 100 and 3 are obtained from the lower quartile statistically calculated from all delivery records in the past seven days. For the time periods where there is an obvious disconnection between clicks and jumps, key attention should be paid to the user browsing duration. When the duration is less than 5 seconds, it is included in the record list of quick exits and the advertisement type and audience characteristics are compared again. At the same time, pay attention to collecting the scrolling operation trajectory and material playback progress of the user after each opening of the advertisement page, associate all browsing actions with the start time of the delivery. If there are repeated access users, identify and merge them for statistics in the background. During the integration, it is also necessary to check whether there is a situation where the continuous access or stay exceeds 60 seconds. This 60-second threshold is set after summarizing the average playback completion rate of video advertisements. If it is found that a certain user has accessed the advertisement page more than 3 times and the access duration is higher than 30 seconds each time, record it as a highly concerned item. After the delivery is completed, summarize the browsing and interaction information of all users on the same day by time period, and combine the actual click quantity and jump depth to form an original data sequence, and finally confirm the exposure volume, click status, and browsing depth of this time period to generate real-time advertisement feedback data.

[0060] The benefit of the formula is to combine the integral value of the user interaction intensity in the time dimension with the difference between the content browsing depth and the page bounce rate, and integrate the average stay duration and the number of user repeated accesses in the denominator part, so that multiple key behavior factors can be balanced and considered under a unified calculation framework.

[0061] The steps for obtaining the parameter are as follows. This parameter represents the user page interaction intensity within the time period. It is necessary to quantify the click actions, page scrolling amplitude, video playback progress, and sliding times generated by the user during the process of browsing the advertisement page and accumulate them within each time segment. If it is divided into 24 time periods in a day, count the interaction actions for each time period, and then combine the counts of all time periods to form an interaction intensity function with time as the horizontal axis. , and then the function is integrated over time to get a total strength value during the monitoring period. For example, during the 24-hour monitoring period of an advertisement, a total of 400 clicks, 320 scrolls, and 120 video starts are recorded. The weighted mapping method is used to set the proportion of clicks, scrolls, and plays. Clicks are mapped to a single contribution value of 1.2, scrolls are mapped to 1.0, and plays are mapped to 0.8. After integration, the interaction value of the period is obtained. This is repeated after traversing all periods to generate , which can be finally obtained in the integral operation For example, the total of all time periods in a day is 600.

[0062] The steps to obtain the parameter are as follows: The content browsing depth is determined by arranging several monitoring points in the advertising page to determine whether the user has browsed to the middle or bottom of the page. Each monitoring point is triggered once it is reached by scrolling, and the value is calculated based on the user's stay time at the text or image position. For example, three monitoring points are set on the page, located at 20%, 50%, and 80% of the page height respectively. When the user triggers each monitoring point in turn, the corresponding depth increment is recorded, and the depth value is accumulated to form a depth value. If the user continues to stay for more than 5 seconds after reaching 80%, an additional depth coefficient of 0.1 is added. The entire measurement process records each visit, and the average browsing depth is obtained after the cumulative visits are completed. For example, the average overall depth of 1,000 users browsing data is 3.2, that is, .

[0063] The steps to obtain the parameter are as follows: The page bounce rate is first collected. The actions of users closing or switching pages within a short period of time after clicking on the advertising page are collected. If the stay time is less than 3 seconds, it is regarded as a quick bounce. The 3-second benchmark value comes from the threshold obtained by classifying 50,000 access records. The number of quick bounces is divided by the total number of visits to get the bounce rate, and an additional mark is added for cases with obvious quick return behavior. For example, the total number of visits in a certain advertising statistical period is 3,000, of which the number of quick bounces reaches 450, then the bounce rate = 450 / 3,000 = 0.15. If it is monitored that some users repeatedly enter and exit quickly, only one bounce is recorded, and finally the bounce rate is used as The value of =0.15.

[0064] The steps to obtain the parameter are as follows: For the average duration of stay in each time period, we first count the number of seconds that users stay on the ad page one by one. When the visit ends, we record the departure time minus the entry time to get the stay interval. We divide the stay time of all users into time periods to get the average value. Finally, we weight and integrate the average values ​​of different time periods to form the overall average value of the entire delivery cycle. For example, the average stay time in the morning, noon and night time periods is calculated as 8 seconds, 10 seconds and 12 seconds respectively, and the weighted average value is obtained by the proportion of visitors in each time period of 0.3, 0.4 and 0.3. In one actual measurement, the value was found to be approximately 10.2 seconds.

[0065] The steps to obtain the parameter are as follows: The number of repeated visits by users is calculated by performing user-level deduplication statistics on all visits to the ad page. If the same user ID is detected to have visited the same ad page multiple times, the number of repeated visits to the ad by the user is accumulated one by one. For example, if a user visits the ad page five times in two days, the number of repeated visits to the ad by the user is recorded as 5. The total number of such visits by all users is added up and then smoothed according to the total number of visitors. If the ad page has a total of 2,000 visits in one cycle, of which 500 are repeated visits, then , the ratio or an integer form can be considered as The final value of .

[0066] The steps to obtain the parameter are as follows: it represents the total duration after the ad is released, that is, the time starts from the time when the ad is released online, and is pushed back in hours or minutes until the release ends or the statistics are closed, and then the duration is recorded as For example, if an ad starts running at 9:00 a.m. and ends at 9:00 the next day, for a total of 24 hours, then Hour.

[0067] Calculation process: make , , , , , first calculate , add the integral value to the difference , then calculate = = = , and Add , and finally: The results show that advertising The delivery result is approximately 57.45. When comparing the values of different advertisements in the same list, the higher the value, the more significant its performance in dimensions such as user interaction intensity, browsing depth, and repeat visits. A too low value indicates that the data indicators in some links are not ideal. At the same time, cross-analysis can be combined with this value and other interaction dimensions obtained previously. If is greater than 50, it means that the overall activity of this advertisement is relatively high during this period. If it is less than 10, it means that the advertisement cannot obtain sufficient user interaction or repeat visits within the available time.

[0068] After calculating the advertisement delivery result, list the values of each advertisement in different time periods, read the visitor information of the advertisement in each time period and compare it with the previously determined comment quantity and page bounce record. If it is found that the value corresponding to a relatively large access volume, then conduct a more in-depth analysis of the browsing depth during this period and check whether the user scrolling trajectory reaches the monitoring points mentioned above. At the same time, pay attention to the stay patterns of repeat visitors and verify whether they are mainly distributed in specific time periods or platforms. Compare all abnormal data information with the threshold range and determine whether any time period with a bounce rate significantly greater than 0.5 is related to content layout or excessive materials. When these time periods match the page layout, make a report comparison. After confirming that all records are associated, generate a time-period delivery comparison table to illustrate the browsing characteristics and key points of concern in each time period. Finally, make a horizontal comparison of this information with the delivery situations of other advertisements to obtain the delivery result.

[0069] The present invention provides an advertisement device, including: a processor and a memory. The memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory, so that the advertisement device executes the above-mentioned advertisement method for device sales based on artificial intelligence.

[0070] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical content of the technical solution of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An advertising method for device sales based on artificial intelligence, characterized in that It includes the following steps: Collect the social media behavior data and purchase records of consumers in the target market, and through data cleaning and formatting processing, obtain standardized consumer data; Based on the standardized consumer data, conduct user portrait analysis, including age, preferences, and purchasing power, to generate consumer segmentation results; Based on the consumer segmentation results, conduct sentiment analysis on the consumer groups to identify the emotional reactions and preferences towards the advertising content, and obtain sentiment preference results; Adjust the advertising copy and visual elements according to the sentiment preference results to generate customized advertising content; Use the customized advertising content to conduct tests on the platform, record the user interaction data and feedback, and obtain the advertising effect evaluation results; based on the advertising effect evaluation results, analyze the performance differences on different platforms, optimize the advertising placement strategy, and generate an optimized advertising deployment plan; According to the optimized advertising deployment plan, conduct advertising releases, and at the same time monitor the real-time data stream, and adjust the advertising content and placement time according to the real-time data stream to obtain the placement results.

2. The advertising method for device sales based on artificial intelligence according to claim 1, wherein The steps for obtaining the standardized consumer data are as follows: Collect the social media behavior data and purchase records of consumers in the target market, remove duplicate data, correct data format errors, and obtain a preliminarily processed data set; Based on the preliminarily processed data set, identify and process missing values, outliers, and data biases, and apply data imputation and outlier correction to obtain standardized consumer data.

3. The advertising method for device sales based on artificial intelligence according to claim 1, wherein The steps for obtaining the consumer segmentation results are as follows: Extract the age, purchase frequency, and preference types of consumers from the standardized consumer data to obtain the feature extraction results; Based on the feature extraction results, calculate the purchasing power index of each consumer, and the calculation formula is: Among them, represents the purchasing power index of consumers , represents the age value of consumers , represents the median of the age values of all consumers represents the purchase frequency of consumers , represents the preference type of consumers , represents the total sum of the purchase frequency values of all consumers in the target market represents the total purchase amount of consumers ; Based on the purchasing power index, combined with the purchase frequency, preference types, and age range of consumers, conduct category attribution judgment on consumers to generate consumer segmentation results.

4. The advertising method for device sales based on artificial intelligence according to claim 1, characterized in that, The steps for obtaining the sentiment preference results are as follows: Extract the advertising interaction data of each group from the consumer segmentation results, including the number of likes, comments, sharing frequency, and advertising viewing duration, conduct data screening and classification, and obtain the consumer group interaction characteristics; Based on the consumer group interaction characteristics, calculate the emotional reaction scores of each group towards the advertising content, and the calculation formula is: Among them, represents the emotional response score of the group , represents the number of likes of the group , represents the number of comments of the group , represents the advertisement sharing frequency of the group , represents the advertisement viewing duration of the group . Based on the emotional reaction scores, combined with the advertising interaction patterns of the consumer groups, analyze the emotional trends of each consumer group towards the advertising content to obtain the sentiment preference results.

5. The advertising method for device sales based on artificial intelligence according to claim 1, wherein The steps for obtaining the customized advertising content are as follows: According to the sentiment preference results, extract the tendencies of different consumer groups towards advertising text and visual design, and generate advertising copy characteristics and visual element characteristics; Based on the advertising copy characteristics and visual element characteristics, screen the advertising text content and visual materials that meet the emotional preferences of the consumer groups, adjust the expression ways of text sentences, advertising theme presentation ways, and information transmission structures, and generate an advertising copy adjustment plan and a visual element adjustment plan; Based on the advertisement copy adjustment plan and the visual element adjustment plan, integrate the optimized copy content and visual materials, adjust the matching degree between the advertisement content and the characteristics of the consumer group, and generate customized advertisement content.

6. The advertising method for device sales based on artificial intelligence according to claim 1, wherein The steps for obtaining the advertisement effect evaluation result are as follows: Use the customized advertisement content to conduct an advertisement placement test on the platform, set advertisement placement strategies, including release time, display frequency, and placement audience range, collect user interaction behavior data, and generate a user interaction data set; Based on the user interaction data set, calculate the advertisement adaptation degree score, and the calculation formula is: Among them, represents the adaptation score of the advertisement , represents the number of clicks on the advertisement , represents the dwell time of the advertisement , represents the number of shares of the advertisement , represents the number of comments on the advertisement , represents the number of impressions of the advertisement . Based on the advertisement adaptation degree score, analyze the performance of the advertisement on the platform, compare the user interaction effects of each advertisement, and obtain the advertisement effect evaluation result.

7. The advertising method for device sales based on artificial intelligence according to claim 1, characterized in that, The steps for obtaining the optimized advertisement deployment plan are as follows: Based on the advertisement effect evaluation result, calculate the performance differences of advertisement placements on each platform, analyze the user behavior patterns of different platforms, compare the effects of advertisement content among different audience groups, and generate an advertisement placement strategy optimization plan; Based on the advertisement placement strategy optimization plan, adjust the advertisement placement time, adjust the display method of the advertisement content on each platform, optimize the adaptation strategy of the advertisement materials, and adjust the budget allocation ratio to form an optimized advertisement deployment plan.

8. The advertising method for device sales based on artificial intelligence according to claim 1, characterized in that The steps for obtaining the placement result are as follows: According to the optimized advertisement deployment plan, conduct advertisement placement, monitor the user behavior data after the advertisement is exposed, and generate advertisement real-time feedback data; Based on the advertisement real-time feedback data, calculate the advertisement placement result, and the expression is: Among them, represents the advertising delivery result, represents the advertising in the time of the user page interaction intensity, represents the advertising content browsing depth, represents the advertising page bounce rate, represents the advertising average stay duration in each time period, represents the advertising number of user repeat visits, T represents the total duration after the advertisement is delivered; Based on the advertisement placement result, analyze the user response situation of the advertisement on each platform and time period to obtain the placement result.

9. Advertising device, characterized in that, Including: A processor and a memory, the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory, so that the advertisement promotion device executes the method according to any one of claims 1 to 8.

Images