Mother and infant crowd emotion content recognition system and method based on AI intellectualization

Abstract

The invention relates to the technical field of user data analysis, in particular to a maternal and infant crowd emotion content recognition system and method based on AI intellectualization, and the method comprises the steps: evaluating the product demands of maternal and infant product purchasers according to child rearing stage data, search behavior data and purchase behavior data; according to the child rearing stage data and the agent dialogue content data, evaluating the emotional expression of the maternal and infant product purchaser during communication with the agent; according to the product demand evaluation result and the emotion expression evaluation result of the maternal and infant product purchaser, analyzing the emotion intimacy of the maternal and infant product purchaser to the intelligent agent; and according to the emotional intimacy of the maternal and infant product purchaser to the intelligent agent, adjusting the dialogue strategy of the intelligent agent. The service strategy of the intelligent agent is adjusted by evaluating the product demand and the emotion expression, so that the dialogue experience between the user and the intelligent agent is improved, a deeper user relationship is established, and the customer loyalty and the overall benefit of the platform are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of user data analysis, and in particular to an AI-based intelligent mother-infant group emotion content recognition system and method. Background Art

[0002] With the continuous development of social economy and the improvement of residents' living standards, the maternal and infant market has shown a rapid growth trend. More and more families have begun to pay attention to the quality and service of maternal and infant products, which has promoted the prosperity of this market. In this context, it is particularly important for enterprises to adjust the interaction between intelligent bodies and maternal and infant populations according to the needs of maternal and infant populations. The needs of maternal and infant populations have significant stage and personalized characteristics. In different stages of parenting, such as pregnancy, newborn, and infant, the needs of maternal and infant populations for products and services are different. For example, pregnant mothers may pay more attention to maternity clothes, nutritional products, and prenatal examination services, while parents in infants and toddlers may pay more attention to baby food, safety products, and early education products. In addition, the emotional state of different stages of parenting is also different. Emotional fluctuations during pregnancy, anxiety of novice parents in the newborn period, and various concerns during the growth of infants and toddlers are all important emotional factors that need to be paid attention to.

[0003] However, the intelligent agents in the existing technology do not take into account the different product demands and emotional expressions of mothers and infants at different parenting stages when interacting with them. Therefore, the dialogue strategy with mothers and infants cannot be accurately adjusted according to the different parenting stages of mothers and infants, resulting in the inability of the intelligent agents to provide truly personalized services, which in turn leads to poor user experience and difficulty in effectively meeting users' personalized needs and emotional demands. Summary of the invention

[0004] In order to overcome the defects and shortcomings of the prior art, the present invention provides an AI-based intelligent mother-infant emotional content recognition system and method, which evaluates product demand and emotional expression and adjusts the service strategy of the intelligent agent, thereby enhancing the dialogue experience between users and the intelligent agent, establishing a deeper user relationship, and improving customer loyalty and the overall benefits of the platform.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: In a first aspect, an embodiment of the present invention provides an AI-based intelligent method for identifying emotional content of mothers and infants, comprising the following steps: S1. Obtain the parenting stage data of maternal and infant product buyers and the conversation content data between the maternal and infant product buyers and the intelligent agent through the intelligent agent, and at the same time obtain the search behavior data and purchase behavior data of the maternal and infant product buyers; S2. Importing the parenting stage data, search behavior data, and purchase behavior data into the maternal and infant product demand assessment model to assess the product demand of maternal and infant product buyers; S3, importing the parenting stage data and the agent conversation content data into the emotion expression evaluation model, and evaluating the emotional expression of maternal and infant product buyers when communicating with the agent; S4. Analyze the emotional intimacy of maternal and infant product buyers towards the intelligent agent based on the product demand evaluation results and emotional expression evaluation results of maternal and infant product buyers; S5. Adjust the dialogue strategy of the intelligent agent based on the emotional intimacy of maternal and infant product buyers towards the intelligent agent.

[0006] In an optional implementation, step S2 includes the following specific steps: S21, inquiring maternal and infant product buyers through an intelligent agent to obtain the parenting stage of the maternal and infant product buyers, and obtaining the search behavior data and purchase behavior data of historical maternal and infant product buyers in the same parenting stage according to the parenting stage of the maternal and infant product buyers; S22. Substituting the search behavior data and purchase behavior data of maternal and infant product buyers and the search behavior data and purchase behavior data of historical maternal and infant product buyers in the same child-rearing stage into the product demand difference coefficient calculation formula to calculate the product demand difference coefficient of maternal and infant product buyers; the product demand difference coefficient calculation formula is: ; Where XC is the product demand difference coefficient of maternal and infant product buyers, Indicates the product price difference index for maternal and infant product buyers, Indicates the product quality difference index of maternal and infant product buyers, and They respectively represent the mean and standard deviation of the product price difference index of all historical maternal and infant product buyers calculated based on the search behavior data and purchase behavior data of all historical maternal and infant product buyers at the same parenting stage as the maternal and infant product buyer; and They respectively represent the mean and standard deviation of the product quality difference index of all historical maternal and infant product buyers calculated based on the search behavior data and purchase behavior data of all historical maternal and infant product buyers in the same parenting stage as the maternal and infant product buyer; tanh() represents the hyperbolic tangent function.

[0007] In an optional implementation, the calculation formula of the product price difference index of the maternal and infant product purchasers is: ; In the formula, represents the product price difference index of maternal and infant product buyers, n is the number of product types purchased by maternal and infant product buyers in the purchase behavior data, mi represents the purchase price of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, N is the number of product types searched by maternal and infant product buyers in the search behavior data, They respectively represent the maximum and minimum values ​​of the price range of all i-th products searched by maternal and infant product buyers in the search behavior data, i is any item from 1 to n, and ln() is a natural logarithm function.

[0008] In an optional implementation, the calculation formula for the product quality difference index of the maternal and infant product purchasers is: ; In the formula, represents the product quality difference index of maternal and infant product buyers, hpi represents the favorable comment rate of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, qi represents the monthly sales volume of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, represents the mean of the favorable reviews of all i-th products searched by maternal and infant product buyers in the search behavior data, It represents the mean monthly sales volume of all i-th products searched by maternal and infant product buyers in the search behavior data, i is any item from 1 to n, and exp() is a natural exponential function.

[0009] In an optional implementation manner, step S3 includes the following specific steps: S31, obtaining conversation content data between the maternal and infant product purchaser and the agent, and according to the parenting stage of the maternal and infant product purchaser, obtaining historical conversation content data between the historical maternal and infant product purchasers and the agent in the same parenting stage; S32, importing the conversation texts input by the maternal and infant product purchasers in the conversation content data into the VADER sentiment analysis tool to perform sentiment analysis on each conversation text input by the maternal and infant product purchasers, obtaining sentiment scores of various types of sentiments in each conversation text input by the maternal and infant product purchasers, and constructing a sentiment score vector for each conversation text input by the maternal and infant product purchasers through the sentiment scores of various types of sentiments, wherein the sentiment score vector of the qth conversation text input by the maternal and infant product purchasers is ; where xf is the sentiment score of all negative sentiment words in the qth dialogue text entered by the maternal and infant product buyer, xz is the sentiment score of all neutral sentiment words in the qth dialogue text entered by the maternal and infant product buyer, xp is the sentiment score of all positive sentiment words in the qth dialogue text entered by the maternal and infant product buyer, and Xq is the composite sentiment score of the qth dialogue text entered by the maternal and infant product buyer, where ; S33, calculating the emotional tendency value em of all dialogue texts input by the maternal and infant product purchasers, the calculation formula of the emotional tendency value is: : Where R is the number of all dialogue texts entered by maternal and infant product buyers, and q is any item from 1 to R.

[0010] In an optional implementation manner, step S3 further includes the following specific steps: S34, substituting the calculated emotion tendency values ​​of all dialogue texts input by maternal and infant product buyers and the emotion tendency values ​​of all dialogue texts input by historical maternal and infant product buyers into the emotion expression difference coefficient calculation formula to calculate the emotion expression difference coefficient of maternal and infant product buyers; the emotion expression difference coefficient calculation formula is: ; In the formula, BC is the difference coefficient of emotional expression of maternal and infant product buyers, W represents the number of historical maternal and infant product buyers at the same parenting stage as the maternal and infant product buyers, is the emotional tendency value of all dialogue texts input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, Tj represents the number of all dialogue texts input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, Xhk represents the emotional composite score of the kth dialogue text input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, k is any item from 1 to Tj, and j is any item from 1 to W.

[0011] In an optional implementation manner, step S4 includes the following specific steps: S41. Obtaining the calculated product demand difference coefficient and emotion expression difference coefficient of maternal and infant product buyers; S42, substituting the product demand difference coefficient and the emotion expression difference coefficient of the maternal and infant product buyers into the emotional intimacy calculation formula to calculate the emotional intimacy of the maternal and infant product buyers to the intelligent agent; the emotional intimacy calculation formula is: ; Where Qm represents the emotional intimacy of maternal and infant product buyers towards the intelligent agent.

[0012] In an optional implementation manner, step S5 includes the following specific steps: S51, obtaining the calculated emotional intimacy between the purchaser of maternal and infant products and the intelligent agent; S52. Preset an emotional intimacy threshold. When the emotional intimacy of the maternal and infant product purchaser to the intelligent agent is greater than the emotional intimacy threshold, maintain the current dialogue strategy between the intelligent agent and the maternal and infant product purchaser; when the emotional intimacy of the maternal and infant product purchaser to the intelligent agent is less than or equal to the emotional intimacy threshold, adjust the dialogue strategy between the intelligent agent and the maternal and infant product purchaser.

[0013] In a second aspect, the embodiment of the present invention also provides an AI-based intelligent mother-infant group emotional content recognition system, including: A data acquisition module is used to acquire the parenting stage data of maternal and infant product buyers and the conversation content data between the maternal and infant product buyers and the intelligent agent through the intelligent agent, and to acquire the search behavior data and purchase behavior data of the maternal and infant product buyers at the same time; The product demand assessment module is used to import parenting stage data, search behavior data, and purchase behavior data into the maternal and infant product demand assessment model to assess the product demand of maternal and infant product buyers; The emotional expression evaluation module is used to import the parenting stage data and the agent dialogue content data into the emotional expression evaluation model to evaluate the emotional expression of maternal and infant product buyers when communicating with the agent; The emotional intimacy analysis module is used to analyze the emotional intimacy of maternal and infant product buyers towards the intelligent agent based on the product demand evaluation results and emotional expression evaluation results of maternal and infant product buyers; The dialogue strategy adjustment module is used to adjust the dialogue strategy of the agent according to the emotional intimacy of the maternal and infant product buyers to the agent; A control module is used to control the operation of the data acquisition module, the product demand assessment module, the emotion expression assessment module, the emotional intimacy analysis module and the dialogue strategy adjustment module.

[0014] In a third aspect, an embodiment of the present invention provides an electronic device, comprising: a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes an AI-based intelligent method for identifying emotional content of mothers and babies by calling the computer program stored in the memory.

[0015] Compared with the prior art, the present invention has the following advantages and beneficial effects: The present invention evaluates the product needs of maternal and infant product buyers based on parenting stage data, search behavior data, and purchase behavior data; evaluates the emotional expression of maternal and infant product buyers when communicating with the intelligent agent based on parenting stage data and intelligent agent dialogue content data; analyzes the emotional intimacy of maternal and infant product buyers to the intelligent agent based on the product demand evaluation results and emotional expression evaluation results of maternal and infant product buyers; and adjusts the dialogue strategy of the intelligent agent based on the emotional intimacy of maternal and infant product buyers to the intelligent agent. By evaluating product needs and emotional expressions and adjusting the service strategy of the intelligent agent, the dialogue experience between users and the intelligent agent is improved, a deeper user relationship is established, and customer loyalty and the overall benefits of the platform are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 It is a schematic diagram of the overall process of the AI ​​intelligent mother-infant group emotional content recognition method of the present invention; Figure 2 The flowchart of the VADER emotion analysis tool in the AI ​​intelligent mother-infant group emotion content recognition method of the present invention is as follows; Figure 3 It is a structural schematic diagram of the AI ​​intelligent mother-infant group emotion content recognition system of the present invention. DETAILED DESCRIPTION

[0017] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. The embodiments of the present invention and the technical features in the embodiments may be combined with each other unless there is a conflict.

[0018] Example 1

[0019] like Figure 1 As shown, this embodiment provides an AI-based intelligent method for identifying emotional content of mothers and infants, which specifically includes the following steps: S1. Obtain the parenting stage data of maternal and infant product buyers and the conversation content data between the maternal and infant product buyers and the intelligent agent through the intelligent agent, and at the same time obtain the search behavior data and purchase behavior data of the maternal and infant product buyers; S2. Importing the parenting stage data, search behavior data, and purchase behavior data into the maternal and infant product demand assessment model to assess the product demand of maternal and infant product buyers; In this embodiment, step S2 includes the following specific steps: S21, inquiring maternal and infant product buyers through an intelligent agent to obtain the parenting stage of the maternal and infant product buyers, and obtaining the search behavior data and purchase behavior data of historical maternal and infant product buyers in the same parenting stage according to the parenting stage of the maternal and infant product buyers; S22. Substituting the search behavior data and purchase behavior data of maternal and infant product buyers and the search behavior data and purchase behavior data of historical maternal and infant product buyers in the same child-rearing stage into the product demand difference coefficient calculation formula to calculate the product demand difference coefficient of maternal and infant product buyers; the product demand difference coefficient calculation formula is: ; Where XC is the product demand difference coefficient of maternal and infant product buyers, Indicates the product price difference index for maternal and infant product buyers, Indicates the product quality difference index of maternal and infant product buyers, and They respectively represent the mean and standard deviation of the product price difference index of all historical maternal and infant product buyers calculated based on the search behavior data and purchase behavior data of all historical maternal and infant product buyers at the same parenting stage as the maternal and infant product buyer; and They respectively represent the mean and standard deviation of the product quality difference index of all historical maternal and infant product buyers calculated based on the search behavior data and purchase behavior data of all historical maternal and infant product buyers in the same parenting stage as the maternal and infant product buyer; tanh() represents the hyperbolic tangent function.

[0020] It should be noted that the hyperbolic tangent function used in this embodiment can smoothly limit the growth of deviation. This means that when the price difference or the quality difference becomes very large, the product demand difference will not increase indefinitely, but tend to a saturation value. This characteristic helps to avoid extreme values ​​from having too much impact on the overall analysis results, thereby improving the stability and robustness of the model. At the same time, the hyperbolic tangent function is a nonlinear function that can introduce nonlinear influencing factors into the model. In economics and statistics, many relationships are not linear, so the use of nonlinear functions can better describe these complex relationships. By introducing the hyperbolic tangent function, the method for calculating product demand differences provided in this embodiment can more accurately capture the impact of price differences and quality differences on product demand differences. Although the hyperbolic tangent function limits the unlimited growth of product demand differences, it remains sensitive to differences. This means that even if the difference is small, the product demand difference coefficient calculation formula provided in this embodiment can capture this change and reflect it in the product demand difference. This sensitivity helps the model to more accurately identify small fluctuations in price or quality, and can avoid the problem of numerical overflow or underflow of the product demand difference coefficient during the calculation process.

[0021] In this embodiment, the calculation formula of the product price difference index of maternal and infant product buyers is: ; In the formula, represents the product price difference index of maternal and infant product buyers, n is the number of product types purchased by maternal and infant product buyers in the purchase behavior data, mi represents the purchase price of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, N is the number of product types searched by maternal and infant product buyers in the search behavior data, They respectively represent the maximum and minimum values ​​of the price range of all i-th products searched by maternal and infant product buyers in the search behavior data, i is any item from 1 to n, and ln() is a natural logarithm function.

[0022] In this embodiment, the calculation formula of the product quality difference index of maternal and infant product buyers is: ; In the formula, represents the product quality difference index of maternal and infant product buyers, hpi represents the favorable comment rate of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, qi represents the monthly sales volume of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, represents the mean of the favorable reviews of all i-th products searched by maternal and infant product buyers in the search behavior data, It represents the mean monthly sales volume of all i-th products searched by maternal and infant product buyers in the search behavior data, i is any item from 1 to n, and exp() is a natural exponential function.

[0023] S3, importing the parenting stage data and the agent conversation content data into the emotion expression evaluation model, and evaluating the emotional expression of maternal and infant product buyers when communicating with the agent; In this embodiment, step S3 includes the following specific steps: S31, obtaining conversation content data between the maternal and infant product purchaser and the agent, and according to the parenting stage of the maternal and infant product purchaser, obtaining historical conversation content data between the historical maternal and infant product purchasers and the agent in the same parenting stage; S32, importing the conversation texts input by the maternal and infant product purchasers in the conversation content data into the VADER sentiment analysis tool to perform sentiment analysis on each conversation text input by the maternal and infant product purchasers, obtaining sentiment scores of various types of sentiments in each conversation text input by the maternal and infant product purchasers, and constructing a sentiment score vector for each conversation text input by the maternal and infant product purchasers through the sentiment scores of various types of sentiments, wherein the sentiment score vector of the qth conversation text input by the maternal and infant product purchasers is ; where xf is the sentiment score of all negative sentiment words in the q-th conversation text input by the mother and baby product purchaser, xz is the sentiment score of all neutral sentiment words in the q-th conversation text input by the mother and baby product purchaser, xp is the sentiment score of all positive sentiment words in the q-th conversation text input by the mother and baby product purchaser, and Xq is the sentiment composite score of the q-th conversation text input by the mother and baby product purchaser, where ; It should be noted that the VADER sentiment analysis tool used in this embodiment is a natural language processing tool designed specifically for text sentiment analysis. The VADER sentiment analysis tool combines a lexicon method and a rule-based system, and can handle special expressions such as abbreviations, capital letters, and exclamation marks commonly found in social media, so as to more accurately capture the text sentiment polarity. The core of the VADER sentiment analysis tool lies in its sentiment lexicon, which contains words with positive and negative sentiment labels and takes into account factors such as word intensity, negative prefixes, and emphasis. In addition, VADER also uses rules and machine learning algorithms to handle complex sentence structures and contexts. This embodiment provides a flowchart of the working process of the VADER sentiment analysis tool, as Figure 2 shown. In this embodiment, the working principle of performing sentiment analysis on the conversation text through the VADER sentiment analysis tool is roughly as follows: A1. Decompose the conversation text input by the mother and baby product purchaser into individual words or phrases; A2. Remove punctuation marks, stop words in the text, such as "de" (of), "le" (already), etc., and irrelevant characters to reduce the impact of noise on sentiment analysis; A3. Restore the words to their basic form, that is, the root form, in order to accurately match the words in the sentiment lexicon; A4. Match the processed words of the conversation text input by the mother and baby product purchaser with the built-in sentiment lexicon of VADER. The sentiment lexicon contains a large number of words with positive and negative sentiment labels and takes into account the intensity of the words; A5. For each matched word, assign a corresponding sentiment score according to its sentiment label and intensity. Among them, the sentiment labels include: positive, negative, neutral; the sentiment intensity includes modifiers such as "very", "slightly", etc.; A6. Identify negative words (such as "bu" (not), "meiyou" (no), etc.) in the conversation text input by the mother and baby product purchaser, and reverse the sentiment polarity of the subsequent words according to the rules. For example, "buhao" (not good) will be recognized as negative sentiment; A7. Adjust the intensity of the sentiment score according to the intensity of the words and the context. For example, "feichanghao" (very good) will have a higher positive sentiment score than "hao" (good); A8. Summarize the sentiment scores of all words in the conversation text input by the maternal and infant product purchaser to obtain the overall positive sentiment word score, negative sentiment word score, and neutral sentiment word score; A9. Determine the composite emotion score of the conversation text input by the maternal and infant product purchaser based on the relative sizes of the positive emotion word scores, the negative emotion word scores, and the neutral emotion word scores.

[0024] In this embodiment, the VADER sentiment analysis tool not only evaluates the overall sentiment of maternal and infant product buyers when communicating with the agent, but also detects subjectivity and objectivity, as well as positive, negative and neutral sentiments, and provides detailed sentiment analysis results. Moreover, the VADER sentiment analysis tool uses a pre-built sentiment dictionary and rule-based system, which can achieve efficient and accurate sentiment analysis without the need for a complex model training process.

[0025] S33, calculating the emotional tendency value em of all dialogue texts input by the maternal and infant product purchasers, the calculation formula of the emotional tendency value is: : Where R is the number of all dialogue texts entered by maternal and infant product buyers, and q is any item from 1 to R.

[0026] In this embodiment, step S3 also includes the following specific steps: S34, substituting the calculated emotion tendency values ​​of all dialogue texts input by maternal and infant product buyers and the emotion tendency values ​​of all dialogue texts input by historical maternal and infant product buyers into the emotion expression difference coefficient calculation formula to calculate the emotion expression difference coefficient of maternal and infant product buyers; the emotion expression difference coefficient calculation formula is: ; In the formula, BC is the difference coefficient of emotional expression of maternal and infant product buyers, W represents the number of historical maternal and infant product buyers at the same parenting stage as the maternal and infant product buyers, is the emotional tendency value of all dialogue texts input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, Tj represents the number of all dialogue texts input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, Xhk represents the emotional composite score of the kth dialogue text input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, k is any item from 1 to Tj, and j is any item from 1 to W.

[0027] S4. Analyze the emotional intimacy of maternal and infant product buyers towards the intelligent agent based on the product demand evaluation results and emotional expression evaluation results of maternal and infant product buyers; In this embodiment, step S4 includes the following specific steps: S41. Obtaining the calculated product demand difference coefficient and emotion expression difference coefficient of maternal and infant product buyers; S42, substituting the product demand difference coefficient and the emotion expression difference coefficient of the maternal and infant product buyers into the emotional intimacy calculation formula to calculate the emotional intimacy of the maternal and infant product buyers to the intelligent agent; the emotional intimacy calculation formula is: ; Where Qm represents the emotional intimacy of maternal and infant product buyers towards the intelligent agent.

[0028] S5. Adjust the dialogue strategy of the intelligent agent based on the emotional intimacy of maternal and infant product buyers towards the intelligent agent.

[0029] In this embodiment, step S5 includes the following specific steps: S51, obtaining the calculated emotional intimacy between the purchaser of maternal and infant products and the intelligent agent; S52. Preset an emotional intimacy threshold. When the emotional intimacy of the maternal and infant product purchaser to the intelligent agent is greater than the emotional intimacy threshold, maintain the current dialogue strategy between the intelligent agent and the maternal and infant product purchaser; when the emotional intimacy of the maternal and infant product purchaser to the intelligent agent is less than or equal to the emotional intimacy threshold, adjust the dialogue strategy between the intelligent agent and the maternal and infant product purchaser, wherein the emotional intimacy threshold is set by technical personnel in this field according to actual conditions and specific needs.

[0030] It should be noted that this embodiment can more accurately understand the preferences and expectations of maternal and infant product buyers for the intelligent agent by accurately calculating emotional intimacy. When emotional intimacy is low, timely adjustment of the dialogue strategy can quickly improve the buyer experience, improve satisfaction, and thus enhance the buyer's loyalty to the brand or product. At the same time, the intelligent agent provides personalized dialogue strategies for maternal and infant product buyers to better meet the personalized needs of buyers, so that buyers feel valued and cared for in their interactions with the intelligent agent, thereby increasing the frequency of interaction between buyers and brands or products and enhancing user stickiness. By adjusting the dialogue strategy according to the buyer's emotional intimacy, the intelligent agent can more effectively guide buyers to complete target behaviors such as purchases and consultations, thereby improving conversion rates and sales.

[0031] Example 2 like Figure 3 As shown, this embodiment provides an AI-based intelligent mother-infant emotional content recognition system, including: A data acquisition module is used to acquire the parenting stage data of maternal and infant product buyers and the conversation content data between the maternal and infant product buyers and the intelligent agent through the intelligent agent, and to acquire the search behavior data and purchase behavior data of the maternal and infant product buyers at the same time; The product demand assessment module is used to import parenting stage data, search behavior data, and purchase behavior data into the maternal and infant product demand assessment model to assess the product demand of maternal and infant product buyers; The emotional expression evaluation module is used to import the parenting stage data and the agent dialogue content data into the emotional expression evaluation model to evaluate the emotional expression of maternal and infant product buyers when communicating with the agent; The emotional intimacy analysis module is used to analyze the emotional intimacy of maternal and infant product buyers towards the intelligent agent based on the product demand evaluation results and emotional expression evaluation results of maternal and infant product buyers; The dialogue strategy adjustment module is used to adjust the dialogue strategy of the agent according to the emotional intimacy of the maternal and infant product buyers to the agent; The control module is used to control the operation of the data acquisition module, the product demand assessment module, the emotion expression assessment module, the emotion intimacy analysis module and the dialogue strategy adjustment module.

[0032] In this embodiment, the product demand evaluation module is used to import the parenting stage data, search behavior data and purchase behavior data into the maternal and infant product demand evaluation model to evaluate the product demand of maternal and infant product buyers. The specific steps of evaluating the product demand of maternal and infant product buyers include: Through the intelligent agent, inquire the maternal and infant product buyers to obtain the parenting stage of the maternal and infant product buyers, and obtain the search behavior data and purchase behavior data of historical maternal and infant product buyers in the same parenting stage according to the parenting stage of the maternal and infant product buyers; Substitute the search behavior data and purchase behavior data of maternal and infant product buyers and the search behavior data and purchase behavior data of historical maternal and infant product buyers in the same child-rearing stage into the product demand difference coefficient calculation formula to calculate the product demand difference coefficient of maternal and infant product buyers; the product demand difference coefficient calculation formula is: ; Where XC is the product demand difference coefficient of maternal and infant product buyers, Indicates the product price difference index for maternal and infant product buyers, Indicates the product quality difference index of maternal and infant product buyers, and They respectively represent the mean and standard deviation of the product price difference index of all historical maternal and infant product buyers calculated based on the search behavior data and purchase behavior data of all historical maternal and infant product buyers at the same parenting stage as the maternal and infant product buyer; and They respectively represent the mean and standard deviation of the product quality difference index of all historical maternal and infant product buyers calculated based on the search behavior data and purchase behavior data of all historical maternal and infant product buyers in the same parenting stage as the maternal and infant product buyer; tanh() represents the hyperbolic tangent function.

[0033] It should be noted that the hyperbolic tangent function used in this embodiment can smoothly limit the growth of deviation. This means that when the price difference or the quality difference becomes very large, the product demand difference will not increase indefinitely, but tend to a saturation value. This characteristic helps to avoid extreme values ​​from having too much impact on the overall analysis results, thereby improving the stability and robustness of the model. At the same time, the hyperbolic tangent function is a nonlinear function that can introduce nonlinear influencing factors into the model. In economics and statistics, many relationships are not linear, so the use of nonlinear functions can better describe these complex relationships. By introducing the hyperbolic tangent function, the method for calculating product demand differences provided in this embodiment can more accurately capture the impact of price differences and quality differences on product demand differences. Although the hyperbolic tangent function limits the unlimited growth of product demand differences, it remains sensitive to differences. This means that even if the difference is small, the product demand difference coefficient calculation formula provided in this embodiment can capture this change and reflect it in the product demand difference. This sensitivity helps the model to more accurately identify small fluctuations in price or quality, and can avoid the problem of numerical overflow or underflow of the product demand difference coefficient during the calculation process.

[0034] In this embodiment, the calculation formula of the product price difference index of maternal and infant product buyers is: ; In the formula, represents the product price difference index of maternal and infant product buyers, n is the number of product types purchased by maternal and infant product buyers in the purchase behavior data, mi represents the purchase price of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, N is the number of product types searched by maternal and infant product buyers in the search behavior data, They respectively represent the maximum and minimum values ​​of the price range of all i-th products searched by maternal and infant product buyers in the search behavior data, i is any item from 1 to n, and ln() is a natural logarithm function.

[0035] In this embodiment, the calculation formula of the product quality difference index of maternal and infant product buyers is: ; In the formula, represents the product quality difference index of maternal and infant product buyers, hpi represents the favorable comment rate of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, qi represents the monthly sales volume of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, represents the mean of the favorable reviews of all i-th products searched by maternal and infant product buyers in the search behavior data, It represents the mean monthly sales volume of all i-th products searched by maternal and infant product buyers in the search behavior data, i is any item from 1 to n, and exp() is a natural exponential function.

[0036] In this embodiment, the emotional expression evaluation module is used to import the parenting stage data and the agent dialogue content data into the emotional expression evaluation model to evaluate the emotional expression of the maternal and infant product purchaser when communicating with the agent. The specific steps of evaluating the emotional expression of the maternal and infant product purchaser when communicating with the agent include: Obtaining conversation content data between maternal and infant product buyers and the agent, and according to the parenting stage of the maternal and infant product buyers, obtaining historical conversation content data between historical maternal and infant product buyers and the agent at the same parenting stage; The conversation texts entered by the maternal and infant product buyers in the conversation content data are imported into the VADER sentiment analysis tool to perform sentiment analysis on each conversation text entered by the maternal and infant product buyers, and the sentiment scores of each type of sentiment in each conversation text entered by the maternal and infant product buyers are obtained. The sentiment score vector of each conversation text entered by the maternal and infant product buyers is constructed through the sentiment scores of each type of sentiment, where the sentiment score vector of the qth conversation text entered by the maternal and infant product buyers is ; where xf is the sentiment score of all negative sentiment words in the qth dialogue text entered by the maternal and infant product buyer, xz is the sentiment score of all neutral sentiment words in the qth dialogue text entered by the maternal and infant product buyer, xp is the sentiment score of all positive sentiment words in the qth dialogue text entered by the maternal and infant product buyer, and Xq is the composite sentiment score of the qth dialogue text entered by the maternal and infant product buyer, where ; It should be noted that the VADER sentiment analysis tool used in this embodiment is a natural language processing tool designed for sentiment analysis of text. The VADER sentiment analysis tool combines a dictionary approach with a rule-based system, and is able to handle special expressions such as abbreviations, capital letters, exclamation marks, etc. commonly used in social media, thereby more accurately capturing the sentiment polarity of the text. The core of the VADER sentiment analysis tool lies in its sentiment dictionary, which contains words with positive and negative sentiment labels, and takes into account factors such as the intensity of words, negative prefixes, and enhanced tone. In addition, VADER also uses rules and machine learning algorithms to handle complex sentence structures and contexts.

[0037] In this embodiment, the VADER sentiment analysis tool not only evaluates the overall sentiment of maternal and infant product buyers when communicating with the agent, but also detects subjectivity and objectivity, as well as positive, negative and neutral sentiments, and provides detailed sentiment analysis results. Moreover, the VADER sentiment analysis tool uses a pre-built sentiment dictionary and rule-based system, which can achieve efficient and accurate sentiment analysis without the need for a complex model training process.

[0038] The emotional tendency value em of all dialogue texts input by maternal and infant product buyers is calculated. The calculation formula of the emotional tendency value is: : Where R is the number of all dialogue texts entered by maternal and infant product buyers, and q is any item from 1 to R.

[0039] In this embodiment, evaluating the emotional expression of the maternal and infant product purchaser when communicating with the intelligent agent also includes the following specific steps: Substitute the calculated emotional tendency values ​​of all dialogue texts input by maternal and infant product buyers and the emotional tendency values ​​of all dialogue texts input by historical maternal and infant product buyers into the emotional expression difference coefficient calculation formula to calculate the emotional expression difference coefficient of maternal and infant product buyers; the emotional expression difference coefficient calculation formula is: ; In the formula, BC is the difference coefficient of emotional expression of maternal and infant product buyers, W represents the number of historical maternal and infant product buyers at the same parenting stage as the maternal and infant product buyers, is the emotional tendency value of all dialogue texts input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, Tj represents the number of all dialogue texts input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, Xhk represents the emotional composite score of the kth dialogue text input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, k is any item from 1 to Tj, and j is any item from 1 to W.

[0040] In this embodiment, the emotional intimacy analysis module is used to analyze the emotional intimacy of the maternal and infant product buyers to the intelligent agent based on the product demand evaluation results and emotional expression evaluation results of the maternal and infant product buyers. The specific steps of analyzing the emotional intimacy of the maternal and infant product buyers to the intelligent agent include: Obtain the calculated product demand difference coefficient and emotion expression difference coefficient of maternal and infant product buyers; Substitute the product demand difference coefficient and the emotion expression difference coefficient of maternal and infant product buyers into the emotional intimacy calculation formula to calculate the emotional intimacy of maternal and infant product buyers to the intelligent agent; the emotional intimacy calculation formula is: ; Where Qm represents the emotional intimacy of maternal and infant product buyers towards the intelligent agent.

[0041] In this embodiment, the dialogue strategy adjustment module is used to adjust the dialogue strategy of the agent according to the emotional intimacy of the maternal and infant product purchaser to the agent. The adjustment of the dialogue strategy of the agent includes the following specific steps: Obtain the calculated emotional intimacy of maternal and infant product buyers towards the agent; A preset emotional intimacy threshold is set. When the emotional intimacy of the maternal and infant product purchaser to the intelligent agent is greater than the emotional intimacy threshold, the current dialogue strategy between the intelligent agent and the maternal and infant product purchaser is maintained; when the emotional intimacy of the maternal and infant product purchaser to the intelligent agent is less than or equal to the emotional intimacy threshold, the dialogue strategy between the intelligent agent and the maternal and infant product purchaser is adjusted, wherein the emotional intimacy threshold is set by technical personnel in this field according to actual conditions and specific needs.

[0042] It should be noted that this embodiment can more accurately understand the preferences and expectations of maternal and infant product buyers for the intelligent agent by accurately calculating emotional intimacy. When emotional intimacy is low, timely adjustment of the dialogue strategy can quickly improve the buyer experience, improve satisfaction, and thus enhance the buyer's loyalty to the brand or product. At the same time, the intelligent agent provides personalized dialogue strategies for maternal and infant product buyers to better meet the personalized needs of buyers, so that buyers feel valued and cared for in their interactions with the intelligent agent, thereby increasing the frequency of interaction between buyers and brands or products and enhancing user stickiness. By adjusting the dialogue strategy according to the buyer's emotional intimacy, the intelligent agent can more effectively guide buyers to complete target behaviors such as purchases and consultations, thereby improving conversion rates and sales.

[0043] The above-mentioned parameters and steps for each unit module to realize the corresponding functions in the AI ​​intelligent mother-infant group emotion content recognition system of the present invention can refer to the parameters and steps in the embodiment of the AI ​​intelligent mother-infant group emotion content recognition method above, and will not be repeated here.

[0044] Example 3 An electronic device according to an embodiment of the present invention comprises: a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes an AI-based intelligent method for identifying emotional content of mothers and infants by calling the computer program stored in the memory. It should be noted that all computer programs of the AI-based intelligent method for identifying emotional content of mothers and infants are implemented in C language, wherein the data acquisition module, product demand assessment module, emotion expression assessment module, emotion intimacy analysis module, dialogue strategy adjustment module and control module are all controlled by a remote server.

[0045] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0046] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0047] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0048] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. The AI-based intelligent method for identifying emotional content of mothers and infants is characterized by: The steps include: S1. Obtain the parenting stage data of maternal and infant product buyers and the conversation content data between the maternal and infant product buyers and the intelligent agent through the intelligent agent, and at the same time obtain the search behavior data and purchase behavior data of the maternal and infant product buyers; S2. Importing the parenting stage data, search behavior data, and purchase behavior data into the maternal and infant product demand assessment model to assess the product demand of maternal and infant product buyers; S3, importing the parenting stage data and the agent conversation content data into the emotion expression evaluation model, and evaluating the emotional expression of maternal and infant product buyers when communicating with the agent; S4. Analyze the emotional intimacy of maternal and infant product buyers towards the intelligent agent based on the product demand evaluation results and emotional expression evaluation results of maternal and infant product buyers; S5. Adjust the dialogue strategy of the intelligent agent based on the emotional intimacy of maternal and infant product buyers towards the intelligent agent.

2. The AI-based intelligent method for identifying emotional content of mothers and infants according to claim 1 is characterized in that: The step S2 comprises the following specific steps: S21, inquiring maternal and infant product buyers through an intelligent agent to obtain the parenting stage of the maternal and infant product buyers, and obtaining the search behavior data and purchase behavior data of historical maternal and infant product buyers in the same parenting stage according to the parenting stage of the maternal and infant product buyers; S22. Substituting the search behavior data and purchase behavior data of maternal and infant product buyers and the search behavior data and purchase behavior data of historical maternal and infant product buyers in the same child-rearing stage into the product demand difference coefficient calculation formula to calculate the product demand difference coefficient of maternal and infant product buyers; the product demand difference coefficient calculation formula is: ; Where XC is the product demand difference coefficient of maternal and infant product buyers, Indicates the product price difference index for maternal and infant product buyers, Indicates the product quality difference index of maternal and infant product buyers, and They respectively represent the mean and standard deviation of the product price difference index of all historical maternal and infant product buyers calculated based on the search behavior data and purchase behavior data of all historical maternal and infant product buyers who are in the same parenting stage as the maternal and infant product buyer; and They respectively represent the mean and standard deviation of the product quality difference index of all historical maternal and infant product buyers calculated based on the search behavior data and purchase behavior data of all historical maternal and infant product buyers in the same parenting stage as the maternal and infant product buyer; tanh() represents the hyperbolic tangent function.

3. The AI-based intelligent method for identifying emotional content of mothers and infants according to claim 2 is characterized in that: The calculation formula of the product price difference index of the maternal and infant product buyers is: ; In the formula, represents the product price difference index of maternal and infant product buyers, n is the number of product types purchased by maternal and infant product buyers in the purchase behavior data, mi represents the purchase price of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, N is the number of product types searched by maternal and infant product buyers in the search behavior data, , They respectively represent the maximum and minimum values ​​of the price range of all i-th products searched by maternal and infant product buyers in the search behavior data, i is any item from 1 to n, and ln() is a natural logarithm function.

4. The AI-based intelligent method for identifying emotional content of mothers and infants according to claim 3 is characterized in that: The calculation formula of the product quality difference index of maternal and infant product buyers is: ; In the formula, represents the product quality difference index of maternal and infant product buyers, hpi represents the favorable comment rate of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, qi represents the monthly sales volume of the i-th product purchased by maternal and infant product buyers in the purchase behavior data, represents the mean of the favorable reviews of all i-th products searched by maternal and infant product buyers in the search behavior data, It represents the mean monthly sales volume of all i-th products searched by maternal and infant product buyers in the search behavior data, i is any item from 1 to n, and exp() is a natural exponential function.

5. The AI-based intelligent method for identifying emotional content of mothers and infants according to claim 4 is characterized in that: The step S3 comprises the following specific steps: S31, obtaining conversation content data between the maternal and infant product purchaser and the agent, and according to the parenting stage of the maternal and infant product purchaser, obtaining historical conversation content data between the historical maternal and infant product purchasers and the agent in the same parenting stage; S32, importing the conversation texts input by the maternal and infant product purchasers in the conversation content data into the VADER sentiment analysis tool to perform sentiment analysis on each conversation text input by the maternal and infant product purchasers, obtaining sentiment scores of various types of sentiments in each conversation text input by the maternal and infant product purchasers, and constructing a sentiment score vector for each conversation text input by the maternal and infant product purchasers through the sentiment scores of various types of sentiments, wherein the sentiment score vector of the qth conversation text input by the maternal and infant product purchasers is ; where xf is the sentiment score of all negative sentiment words in the qth dialogue text entered by the maternal and infant product buyer, xz is the sentiment score of all neutral sentiment words in the qth dialogue text entered by the maternal and infant product buyer, xp is the sentiment score of all positive sentiment words in the qth dialogue text entered by the maternal and infant product buyer, and Xq is the composite sentiment score of the qth dialogue text entered by the maternal and infant product buyer, where ; S33, calculating the emotional tendency value em of all dialogue texts input by the maternal and infant product purchasers, the calculation formula of the emotional tendency value is: ： Where R is the number of all dialogue texts entered by maternal and infant product buyers, and q is any item from 1 to R.

6. The AI-based intelligent method for identifying emotional content of mothers and infants according to claim 5 is characterized in that: The step S3 also includes the following specific steps: S34, substituting the calculated emotion tendency values ​​of all dialogue texts input by maternal and infant product buyers and the emotion tendency values ​​of all dialogue texts input by historical maternal and infant product buyers into the emotion expression difference coefficient calculation formula to calculate the emotion expression difference coefficient of maternal and infant product buyers; the emotion expression difference coefficient calculation formula is: ； In the formula, BC is the difference coefficient of emotional expression of maternal and infant product buyers, W represents the number of historical maternal and infant product buyers at the same parenting stage as the maternal and infant product buyers, is the emotional tendency value of all dialogue texts input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, Tj represents the number of all dialogue texts input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, Xhk represents the emotional composite score of the kth dialogue text input by the jth historical maternal and infant product purchaser who is at the same parenting stage as the maternal and infant product purchaser, k is any item from 1 to Tj, and j is any item from 1 to W.

7. The AI-based intelligent method for identifying emotional content of mothers and infants according to claim 6 is characterized in that: The step S4 comprises the following specific steps: S41. Obtaining the calculated product demand difference coefficient and emotion expression difference coefficient of maternal and infant product buyers; S42, substituting the product demand difference coefficient and the emotion expression difference coefficient of the maternal and infant product buyers into the emotional intimacy calculation formula to calculate the emotional intimacy of the maternal and infant product buyers to the intelligent agent; the emotional intimacy calculation formula is: ； Where Qm represents the emotional intimacy of maternal and infant product buyers towards the intelligent agent.

8. The AI-based intelligent method for identifying emotional content of mothers and infants according to claim 7 is characterized in that: The step S5 comprises the following specific steps: S51, obtaining the calculated emotional intimacy between the purchaser of maternal and infant products and the intelligent agent; S52. Preset an emotional intimacy threshold. When the emotional intimacy of the maternal and infant product purchaser to the intelligent agent is greater than the emotional intimacy threshold, maintain the current dialogue strategy between the intelligent agent and the maternal and infant product purchaser; when the emotional intimacy of the maternal and infant product purchaser to the intelligent agent is less than or equal to the emotional intimacy threshold, adjust the dialogue strategy between the intelligent agent and the maternal and infant product purchaser.

9. An AI-based intelligent mother-infant emotional content recognition system, which is implemented based on the AI-based intelligent mother-infant emotional content recognition method according to any one of claims 1 to 8, characterized in that: The system comprises: A data acquisition module is used to acquire the parenting stage data of maternal and infant product buyers and the conversation content data between the maternal and infant product buyers and the intelligent agent through the intelligent agent, and to acquire the search behavior data and purchase behavior data of the maternal and infant product buyers at the same time; The product demand assessment module is used to import parenting stage data, search behavior data, and purchase behavior data into the maternal and infant product demand assessment model to assess the product demand of maternal and infant product buyers; The emotional expression evaluation module is used to import the parenting stage data and the agent dialogue content data into the emotional expression evaluation model to evaluate the emotional expression of maternal and infant product buyers when communicating with the agent; The emotional intimacy analysis module is used to analyze the emotional intimacy of maternal and infant product buyers towards the intelligent agent based on the product demand evaluation results and emotional expression evaluation results of maternal and infant product buyers; The dialogue strategy adjustment module is used to adjust the dialogue strategy of the agent according to the emotional intimacy of the maternal and infant product buyers to the agent; A control module is used to control the operation of the data acquisition module, the product demand assessment module, the emotion expression assessment module, the emotional intimacy analysis module and the dialogue strategy adjustment module.

10. An electronic device comprising: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; characterized in that the processor executes the AI ​​intelligent mother-infant group emotional content recognition method as described in any one of claims 1-8 by calling the computer program stored in the memory.

Images