User Satisfaction Simulation Method and System Based on Convolutional Neural Network

Abstract

The present invention discloses a user satisfaction simulation method and system based on a convolutional neural network, which includes collecting the behavior data and product test data of users, and preprocessing the behavior data and product test data; using the convolutional neural network to extract the characteristic rules of the behavior data to obtain characteristic data, and performing behavior semantic recognition on the characteristic data to obtain semantic data; correlating the product test data with the semantic data to obtain correlation data, performing directed weighting on the correlation data to obtain data weights, and constructing a satisfaction function using the semantic data and the data weights; constructing a user satisfaction simulation model according to the satisfaction function, optimizing the user satisfaction simulation model, inputting the data to be simulated into the user satisfaction simulation model, and outputting the simulation result. This method not only improves the accuracy of user satisfaction simulation, but also has good interpretability and can be directly applied to the user satisfaction simulation system.

Description

Technical Field

[0001] The present invention relates to the field of simulation, and particularly to a user satisfaction simulation method and system based on a convolutional neural network. Background Art

[0002] In today's digital age, user satisfaction has become one of the key indicators for measuring the success of products or services. To more accurately understand and predict users' satisfaction with products, enterprises and service providers increasingly rely on big data analysis and machine learning technologies. Traditional user satisfaction assessment methods, such as questionnaires and user interviews, although able to provide certain feedback, are often limited by issues such as sample size, strong subjectivity, and lagging feedback. Therefore, it is particularly important to develop a simulation method that can automatically, efficiently, and accurately evaluate user satisfaction.

[0003] With the rapid development of artificial intelligence technology, especially the wide application of convolutional neural networks in the field of deep learning in image recognition, natural language processing, etc., it provides new ideas for user satisfaction assessment. With its powerful feature extraction ability, the convolutional neural network can automatically learn and extract key features from massive and complex data, which is crucial for analyzing the implicit patterns in user behavior data.

[0004] However, directly applying CNN to user satisfaction assessment still faces many challenges. Therefore, there is an urgent need for a new user satisfaction simulation method based on convolutional neural networks to improve the accuracy and real-time performance of satisfaction simulation. Summary of the Invention

[0005] The purpose of the present invention is to provide a user satisfaction simulation method based on a convolutional neural network.

[0006] To achieve the above object, the present invention is implemented according to the following technical solution:

[0007] The present invention includes the following steps:

[0008] Collect users' behavior data and product test data, and preprocess the behavior data and the product test data;

[0009] Use a convolutional neural network to extract characteristic rules from the behavior data to obtain characteristic data, and perform behavior semantic recognition on the characteristic data to obtain semantic data;

[0010] Associate the product test data with the semantic data to obtain associated data, perform directed weighting on the associated data to obtain data weights, and construct a satisfaction function using the semantic data and the data weights;

[0011] Construct the user satisfaction simulation model according to the satisfaction function, optimize the user satisfaction simulation model, input the data to be simulated into the user satisfaction simulation model, and output the simulation results.

[0012] Further, a method for extracting characteristic rules from the behavior data by using a convolutional neural network includes:

[0013] Use a convolutional neural network to extract the image features of the behavior data to obtain image features, and screen the image features. The expression is:

[0014] ,

[0015] ,

[0016] ,

[0017] where the first image feature is , the second image feature is , the binary exclusive OR operation is , the function for counting the total number of bits with the statistical operation result of 1 is , the Hamming distance between the first image feature and the second image feature is H, and the minimum value of the Hamming distance is , the remainder is %, and the maximum value of the Hamming distance is , the Hamming distance between the bth pair of image features is , the automatically varying parameter is , and the set constants are respectively 、 ;

[0018] Further screen the image features that meet the Hamming distance judgment condition. The expression is:

[0019] ,

[0020] ,

[0021] where the window centered on the image feature is G, the offset coordinates inside the window G are , the first image feature at the s+1th moment is , the first image feature at the sth moment is , the similarity score between the first image feature and is U, the first image feature of the behavior data at the s+1th moment is , the first image feature of the behavior data at the sth moment is , and the similarity of the first image features of the first behavior data and the second behavior data is ;

[0022] Given a similarity threshold, when the similarity U is less than the similarity threshold, the image features are output as key features;

[0023] Calculate the change degree of the behavior data:

[0024] ,

[0025] where the b-th key feature at the (s + 1)-th moment is , the b-th key feature at the s-th moment is , the number of key features is , the control optimization coefficient is , the Euclidean norm is , the adjustment parameter is , the similarity of the b-th key feature between the s-th moment and the (s + 1)-th moment is , the change degree of the behavior data between the s-th moment and the (s + 1)-th moment is ;

[0026] Classify the change degree of the behavior data by group and output it as characteristic data.

[0027] Furthermore, a method for obtaining semantic data by performing behavior semantic recognition on the characteristic data includes:

[0028] Input the characteristic data into a behavior semantic recognition model, which includes a non-local feature extraction module and an attention mechanism module;

[0029] The non-local feature extraction module reorganizes the semantic information matrix fused in the dynamic range through reshaping, and makes the local semantic information and the global semantic information interact through matrix multiplication to obtain behavior information;

[0030] Calculate the similarity between the behavior information and the characteristic data:

[0031] ,

[0032] where the query vector is Q, the key vector is K, the value vector is V, the dimension of K is , the transpose is T, the semantic information is Z, the weight matrix of the query vector is , the weight matrix of the key vector is , the weight matrix of the value vector is , the regulation coefficient is , the characteristic data is u, the behavior information is f, and the bias is y;

[0033] The attention mechanism module performs spatio-temporal modeling on the behavior information through a graph convolutional network that utilizes information interaction, and introduces an adaptive parameter, and the expression is:

[0034] ,

[0035] where the adaptive parameter is , the channel dimension is D, and the constants are p and v respectively;

[0036] Adjust the channel weights of the feature information according to the importance of the graph convolutional network through the adaptive parameter, aggregate and extract the effective behavior information of the feature data, and calculate the importance of the feature data:

[0037] ,

[0038] ,

[0039] where the a-th feature data is , the feature data has an importance of , the number of behavior information is , the c-th behavior information is , the j-th behavior information is , the number of feature data is M, and the similarity between the behavior information and the behavior information in the feature data is , and the norm is ;

[0040] Perform clustering division according to the similarity between behavior information to obtain clustering data, and output the clustering data as semantic data.

[0041] Furthermore, the method of constructing a satisfaction function using the semantic data and the data weights includes:

[0042] Classify the semantic data through natural language semantic analysis of Naive Bayes, and convert the semantic data into attitude score data;

[0043] Construct a satisfaction function according to the attitude score data and the data weights, and the expression is:

[0044] ,

[0045] where the s-th data weight is , the c-th attitude score data of the s-th semantic data is , the number of semantic data is , the number of semantic data is , and the satisfaction function is .

[0046] Further, the method for constructing the user satisfaction simulation model according to the satisfaction function includes:

[0047] Construct the objective function of the user satisfaction simulation model by adding the satisfaction function and the loss function;

[0048] The user satisfaction simulation model includes an image feature extraction algorithm, a feature comparison algorithm, a convolutional neural network, and a particle swarm optimization algorithm;

[0049] The image feature extraction algorithm extracts features from the input image data to obtain user features;

[0050] The feature comparison algorithm compares the user features with the behavioral semantic features to obtain comparison data;

[0051] The convolutional neural network mines the relevance based on the comparison data, combines the historical behavior and preferences of the user based on the relevance, and outputs the user simulation satisfaction using the target data;

[0052] The particle swarm optimization algorithm continuously iterates through the variance between the user simulation satisfaction and the actual user satisfaction until the variance is less than 0.13, at which point the iteration stops.

[0053] Further, the method for optimizing the user satisfaction simulation model includes:

[0054] Introduce the intelligent swarm optimization algorithm, use the satisfaction error as the fitness function, and take the minimum satisfaction error as the search target;

[0055] Perform a chaotic mapping on the particle population, and the expression is:

[0056] ,

[0057] where the i-th original sequence is , and the (i + 1)-th chaotic mapping sequence is , the control parameter is , the random number from 0 to 1 is , and the convergence parameter is ;

[0058] Take the position of the particle with the minimum fitness as the best position, and calculate the position of the particle:

[0059] ,

[0060] where the initial best position of the particle at the t-th iteration is , the convergence parameter is , the position of the w-th particle at the (t + 1)-th iteration is , the random number between -1 and 1 is , and the position of the w-th particle at the t-th iteration is ;

[0061] The position of the particle is updated using the suppression factor to obtain the suppression position, and the expression is:

[0062] ,

[0063] ,

[0064] where the suppression position of the w-th particle at the (t + 1)-th iteration is , the suppression factor is , the maximum number of iterations is , the current iteration number is t, the random numbers between 0 and 1 are respectively 、 , the position of the random particle at the t-th iteration is ;

[0065] The position of the particle is updated using the dynamic selection probability to obtain the dynamic position, and the expression is:

[0066] ,

[0067] ,

[0068] where the adaptive weight at the t-th iteration is , the suppression position of the w-th particle at the t-th iteration is , the current iteration number is t, the maximum number of iterations is , the random number of the normal distribution is , the best position at the t-th iteration is , the dynamic position of the w-th particle at the (t + 1)-th iteration is , the dynamic selection probability is ;

[0069] Iterate continuously until the satisfaction error reaches the minimum, otherwise adjust the particle population and update the dynamic selection probability.

[0070] In the second aspect, a user satisfaction simulation system based on a convolutional neural network includes:

[0071] Data acquisition module: used to collect the behavior data and product test data of users, and preprocess the behavior data and the product test data;

[0072] Feature extraction module: used to extract the characteristic rules of the behavior data using a convolutional neural network to obtain characteristic data, and perform behavior semantic recognition on the characteristic data to obtain semantic data;

[0073] Weighted construction module: used to associate the product test data with the semantic data to obtain associated data, perform directed weighting on the associated data to obtain data weights, and construct a satisfaction function using the semantic data and the data weights;

[0074] Model construction module: used to construct the user satisfaction simulation model according to the satisfaction function, optimize the user satisfaction simulation model, input the data to be simulated into the user satisfaction simulation model, and output the simulation result.

[0075] The beneficial effects of the present invention are:

[0076] The present invention is a user satisfaction simulation method and system based on a convolutional neural network. Compared with the prior art, the present invention has the following technical effects:

[0077] Through the steps of preprocessing, characteristic law extraction, behavior semantic recognition, data association, directed weighting, construction of a satisfaction function, model construction, and model optimization, the present invention can improve the accuracy of user satisfaction simulation based on a convolutional neural network, thereby improving the precision of user satisfaction simulation based on a convolutional neural network. Optimizing the user satisfaction simulation based on a convolutional neural network can greatly save resources and improve work efficiency. It can realize intelligent simulation of user satisfaction, extract characteristic laws and recognize behavior semantics in real time for user satisfaction simulation based on a convolutional neural network, which is of great significance for user satisfaction simulation based on a convolutional neural network. It can adapt to user satisfaction simulations based on convolutional neural networks with different standards and different requirements for user satisfaction simulations based on convolutional neural networks, and has a certain universality. Description of the Drawings

[0078] Figure 1 It is a flowchart of the steps of the user satisfaction simulation method based on a convolutional neural network of the present invention. Detailed Embodiments

[0079] The present invention will be further described below through specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention, but do not limit the present invention.

[0080] The user satisfaction simulation method and system based on a convolutional neural network of the present invention include the following steps:

[0081] As Figure 1 shown, in this embodiment, it includes the following steps:

[0082] Collect the behavior data and product test data of users, and preprocess the behavior data and the product test data;

[0083] In the actual evaluation, the behavior data of user A using the live courses on the XXX online education system and the product test data of the XXX online education system are collected;

[0084] The course ID is 4001, the duration is 30 minutes, the difficulty level is medium, and the teacher's score is 4.82;

[0085] The convolutional neural network is used to extract the characteristic rules of the behavior data to obtain characteristic data, and the behavior semantics of the characteristic data is recognized to obtain semantic data;

[0086] In the actual evaluation, the characteristic data of user A is disappearing for 1 minute, the corners of the mouth rising 2 times, opening the mouth wide 3 times, and the attention concentration duration is 26 minutes;

[0087] The semantic data is that there are some difficulties in the course, the overall feeling of the course is good, and the teacher's explanation is good;

[0088] The product test data is associated with the semantic data to obtain associated data, the data weight is obtained by directed weighting according to the associated data, and the satisfaction function is constructed by using the semantic data and the data weight;

[0089] In the actual evaluation, the weighted association rule mining algorithm is used to perform directed weighting on the semantic data to obtain the primary data weight, and the product test data is used to correct the primary data weight to obtain the data weight;

[0090] The data weights are 0.258 for there are some difficulties in the course, 0.896 for the overall feeling of the course is good, and 0.831 for the teacher's explanation is good;

[0091] The user satisfaction simulation model is constructed according to the satisfaction function, the user satisfaction simulation model is optimized, and the data to be simulated is input into the user satisfaction simulation model to output the simulation result.

[0092] In this embodiment, the method for using the convolutional neural network to extract the characteristic rules of the behavior data to obtain the characteristic data includes:

[0093] The convolutional neural network is used to extract the image features of the behavior data to obtain image features, and the image features are screened. The expression is:

[0094] ,

[0095] ,

[0096] ,

[0097] where the first image feature is , and the second image feature is , the binary exclusive OR operation is , the function for counting the total number of bits with the operation result of 1 is , the Hamming distance between the first image feature and the second image feature is H, and the minimum value of the Hamming distance is , the remainder is %, and the maximum value of the Hamming distance is , the Hamming distance between the b-th pair of image features is , the automatically varying parameter is , the set constants are respectively 、 ;

[0098] Further screen the image features that meet the Hamming distance determination condition, and the expression is:

[0099] ,

[0100] ,

[0101] where the window centered on the image feature is G, and the offset coordinates inside the window G are , the first image feature at the s + 1-th moment is , the first image feature at the s-th moment is , the first image feature and The similarity score is U. The first image feature of the behavior data at the s + 1-th moment is , the first image feature of the behavior data at the s-th moment is , the similarity of the first image features between the first behavior data and the second behavior data is ;

[0102] Given a similarity threshold, when the similarity U is less than the similarity threshold, output the image feature as a key feature;

[0103] Calculate the change degree of the behavior data:

[0104] ,

[0105] where the b-th key feature at the s + 1-th moment is , the b-th key feature at the s-th moment is , the number of key features is , the control optimization coefficient is , the Euclidean norm is , the adjustment parameter is , the similarity of the b-th key feature between the s-th moment and the s + 1-th moment is , the change degree of the behavior data between the s-th moment and the s + 1-th moment is ;

[0106] Classify the change degree of behavior data by group and output it as characteristic data.

[0107] In this embodiment, the method for obtaining semantic data by performing behavioral semantic recognition on the characteristic data includes:

[0108] Input the characteristic data into a behavioral semantic recognition model, which includes a non-local feature extraction module and an attention mechanism module;

[0109] The non-local feature extraction module reorganizes the semantic information matrix fused in the dynamic range through reshaping, and makes the local semantic information interact with the global semantic information through matrix multiplication to obtain behavioral information;

[0110] Calculate the similarity between the behavioral information and the characteristic data:

[0111] ,

[0112] where the query vector is Q, the key vector is K, the value vector is V, the dimension of K is , transposed to T, the semantic information is Z, the weight matrix of the query vector is , the weight matrix of the key vector is , the weight matrix of the value vector is , the regulation coefficient is , the characteristic data is u, the behavioral information is f, and the bias is y;

[0113] The attention mechanism module performs spatio-temporal modeling on the behavioral information by using a graph convolutional network for information interaction, introducing an adaptive parameter, and the expression is:

[0114] ,

[0115] where the adaptive parameter is , the channel dimension is D, and the constants are p and v respectively;

[0116] Adjust the channel weights of the feature information according to the importance of the graph convolutional network through the adaptive parameter, aggregate and extract the effective behavioral information of the characteristic data, and calculate the importance of the characteristic data:

[0117] ,

[0118] ,

[0119] where the a-th characteristic data is , the characteristic data has an importance of , the number of behavioral information is , and the c-th behavioral information is The j-th row behavior information is , the number of characteristic data is M, and the characteristic data in the behavior information and the behavior information has a similarity of , and the norm is ;

[0120] Cluster and partition according to the similarity between behavior information to obtain cluster data, and output the cluster data as semantic data.

[0121] In this embodiment, the method of constructing a satisfaction function using the semantic data and the data weight includes:

[0122] Classify the semantic data through natural language semantic analysis of Naive Bayes, and convert the semantic data into attitude score data;

[0123] Construct a satisfaction function according to the attitude score data and the data weight, and the expression is:

[0124] ,

[0125] where the s-th data weight is , the c-th attitude score data of the s-th semantic data is , the number of semantic data is , the number of semantic data is , and the satisfaction function is .

[0126] In this embodiment, the method of constructing the user satisfaction simulation model according to the satisfaction function includes:

[0127] Construct the objective function of the user satisfaction simulation model by adding the satisfaction function and the loss function;

[0128] The user satisfaction simulation model includes an image feature extraction algorithm, a feature comparison algorithm, a convolutional neural network, and a particle swarm optimization algorithm;

[0129] The image feature extraction algorithm extracts features from the input image data to obtain user features;

[0130] The feature comparison algorithm compares the user features with the behavior semantic features to obtain comparison data;

[0131] The convolutional neural network mines the relevance according to the comparison data, combines the historical behavior and preferences of the user based on the relevance, and outputs the user simulation satisfaction using the target data;

[0132] The particle swarm optimization algorithm iterates continuously through the variance between the user simulation satisfaction and the actual user satisfaction until the variance is below 0.13, at which point the iteration stops.

[0133] In this embodiment, the method for optimizing the user satisfaction simulation model includes:

[0134] Introduce the intelligent swarm optimization algorithm, use the satisfaction error as the fitness function, and take the minimum satisfaction error as the search target;

[0135] Perform a chaotic mapping on the particle population, and the expression is:

[0136] ,

[0137] where the i-th original sequence is ,and the (i + 1)-th chaotic mapping sequence is ,the control parameter is ,the random number from 0 to 1 is ,and the convergence parameter is ;

[0138] Take the position of the particle with the minimum fitness as the best position, and calculate the position of the particle:

[0139] ,

[0140] where the initial best position of the particle at the t-th iteration is ,the convergence parameter is ,the position of the w-th particle at the (t + 1)-th iteration is ,the random number between -1 and 1 is ,the position of the w-th particle at the t-th iteration is ;

[0141] Update the position of the particle using the suppression factor to obtain the suppression position, and the expression is:

[0142] ,

[0143] ,

[0144] where the suppression position of the w-th particle at the (t + 1)-th iteration is ,the suppression factor is ,the maximum number of iterations is ,the current number of iterations is t, and the random numbers from 0 to 1 are respectively 、 ,the position of the random particle at the t-th iteration is ;

[0145] Update the position of the particle using the dynamic selection probability to obtain the dynamic position, and the expression is:

[0146] ,

[0147] ,

[0148] where the adaptive weight of the t-th iteration is , the suppression position of the w-th particle in the t-th iteration is , the current iteration number is t, and the maximum iteration number is , the normally distributed random number is , the best position of the t-th iteration is , the dynamic position of the w-th particle in the (t + 1)-th iteration is , the dynamic selection probability is ;

[0149] Iterate continuously until the satisfaction error reaches the minimum, otherwise adjust the particle population and update the dynamic selection probability.

[0150] In a second aspect, a user satisfaction simulation system based on a convolutional neural network includes:

[0151] Data acquisition module: used to collect the behavior data of users and product test data, and preprocess the behavior data and the product test data;

[0152] Feature extraction module: used to extract the characteristic rules of the behavior data by using a convolutional neural network to obtain characteristic data, and perform behavior semantic recognition on the characteristic data to obtain semantic data;

[0153] Weighted construction module: used to associate the product test data with the semantic data to obtain associated data, perform directed weighting according to the associated data to obtain data weights, and construct a satisfaction function by using the semantic data and the data weights;

[0154] Model construction module: used to construct the user satisfaction simulation model according to the satisfaction function, optimize the user satisfaction simulation model, input the data to be simulated into the user satisfaction simulation model, and output the simulation result.

[0155] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A user satisfaction simulation method based on convolutional neural network, characterized in that: The following steps are involved: Collecting user behavior data and product test data, and preprocessing the behavior data and the product test data; Using a convolutional neural network to extract characteristic rules from the behavior data to obtain characteristic data, and performing behavioral semantic recognition on the characteristic data to obtain semantic data; including: The image features of the behavior data are extracted using a convolutional neural network to obtain the image features, and the image features are screened. The expression is: ， ， ， The first image feature is , the second image feature is , the binary XOR operation algorithm is , the total number of digits whose statistical operation result is 1 is , the Hamming distance between the first image feature and the second image feature is H, and the minimum value of the Hamming distance is , the remainder is %, and the maximum value of the Hamming distance is , the Hamming distance between the b-th image feature pair is , the automatic change parameters are , set the constants to , ; To satisfy the Hamming distance The image features of the judgment conditions are further screened, and the expression is: ， ， The window centered on the image feature is G, and the offset coordinates inside the window G are , the first image feature at the s+1th moment is , the first image feature at the sth moment is , the first image feature and The similarity score is U, and the first image feature of the behavior data at the s+1th moment is , the first image feature of the behavior data at the sth moment is , the similarity of the first image feature of the first behavior data and the second behavior data is ; Given a similarity threshold, when the similarity U is less than the similarity threshold, the image feature is output as the key feature; Calculate the degree of change of behavioral data: ， The bth key feature at the s+1th moment is , the bth key feature at the sth moment is , the number of key features is , the control coefficient is , the Euclidean norm is , adjust the parameters to , the similarity of the bth key feature between the sth moment and the s+1th moment is , the change degree of the bth behavior data between the sth moment and the s+1th moment is ; Classify the degree of change of behavioral data into groups and output them as characteristic data; Associating the product test data with the semantic data to obtain associated data, performing directed weighting according to the associated data to obtain data weights, and constructing a satisfaction function using the semantic data and the data weights; A user satisfaction simulation model is constructed according to the satisfaction function, the user satisfaction simulation model is optimized, the data to be simulated is input into the user satisfaction simulation model, and the simulation result is output.

2. The user satisfaction simulation method based on convolutional neural network according to claim 1 is characterized in that: The method of performing behavioral semantic recognition on the characteristic data to obtain semantic data includes: The characteristic data is input into a behavior semantic recognition model, which includes a non-local feature extraction module and an attention mechanism module; The non-local feature extraction module reorganizes the semantic information matrix obtained by fusion in the dynamic range by reshaping, and obtains behavioral information by enabling local semantic information to interact with global semantic information through matrix multiplication; Calculate the similarity between behavior information and feature data: ， The query vector is Q, the key vector is K, the value vector is V, and the dimension of K is , transposed to T, semantic information is Z, and the weight matrix of the query vector is , the weight matrix of the key vector is , the weight matrix of the value vector is , the control coefficient is , the characteristic data is u, the behavior information is f, the bias is y, and the similarity between the characteristic data u and the behavior information f is ; The attention mechanism module uses the graph convolutional network of information interaction to model the behavior information in time and space, and introduces adaptive parameters, which are expressed as: ， The adaptive parameters are , the channel dimension is D, and the constants are p and v respectively; The channel weights of feature information are adjusted according to the importance of the graph convolutional network through adaptive parameters, and the effective behavior information of feature data is extracted through aggregation to calculate the importance of feature data: ， ， The ath characteristic data is , characteristic data The importance of , the amount of behavior information is , the cth behavior information is , the jth behavior information is , the number of characteristic data is M, characteristic data China Bank Information and behavioral information The similarity is , the norm is ; Clustering is performed according to the similarity between behavioral information to obtain cluster data, and the cluster data is output as semantic data.

3. The user satisfaction simulation method based on convolutional neural network according to claim 1 is characterized in that: The method of constructing a satisfaction function using the semantic data and the data weight comprises: Classify semantic data through natural language semantic analysis of Naive Bayes and convert semantic data into attitude score data; The satisfaction function is constructed based on the attitude score data and data weights. The expression is: ， Among them The data weight is , No. The cth attitude score data of semantic data is , the number of attitude score data is , the amount of semantic data is , the satisfaction function is , the average value of attitude score data is .

4. The user satisfaction simulation method based on convolutional neural network according to claim 1 is characterized in that: The method for constructing a user satisfaction simulation model according to the satisfaction function comprises: The objective function of the user satisfaction simulation model is constructed by adding the satisfaction function and the loss function; The user satisfaction simulation model includes image feature extraction algorithm, feature comparison algorithm, convolutional neural network, and particle swarm optimization algorithm; The image feature extraction algorithm extracts features from the input image data to obtain user features; The feature comparison algorithm compares user features with behavioral semantic features to obtain comparison data; The convolutional neural network mines correlations based on the comparative data, combines the user's historical behavior and preferences based on the correlation, and uses the target data to output the user's simulated satisfaction; The particle swarm optimization algorithm continuously iterates through the variance of user simulation satisfaction and user actual satisfaction until the variance is lower than 0.13, then stops iterating.

5. The user satisfaction simulation method based on convolutional neural network according to claim 1 is characterized in that: The method for optimizing the user satisfaction simulation model comprises: The intelligent group optimization algorithm is introduced, the satisfaction error is used as the fitness function, and the minimum satisfaction error is used as the search target; The chaotic mapping of particle population is expressed as: ， The i-th original sequence is , the i+1th chaotic mapping sequence is , the control parameters are , a random number from 0 to 1 is , the convergence parameter is ; Take the particle position with the smallest fitness as the best position and calculate the particle position: ， The initial optimal position of the particle in the tth iteration is , the convergence parameter is , the position of the wth particle in the t+1th iteration is , a random number between -1 and 1 is , the position of the wth particle in the tth iteration is ; Use the inhibition factor to update the particle position and obtain the inhibition position. The expression is: ， ， The suppression position of the wth particle in the t+1th iteration is The inhibitory factor is , the maximum number of iterations is , the current iteration number is t, and the random numbers from 0 to 1 are , , the position of the random particle at the tth iteration is ; The dynamic selection probability is used to update the position of the particle to obtain the dynamic position, which is expressed as: ， ， The adaptive weight of the tth iteration is , the suppression position of the wth particle in the tth iteration is , the current number of iterations is t, and the maximum number of iterations is , the normal distribution random number is , the best position of the tth iteration is , the dynamic position of the wth particle in the t+1th iteration is , the dynamic selection probability is ; Iterate continuously until the satisfaction error is minimized, otherwise adjust the particle population and update the dynamic selection probability.

6. A user satisfaction simulation system based on convolutional neural network, used to execute the method according to any one of claims 1 to 5, characterized in that: include: Data collection module: used to collect user behavior data and product test data, and pre-process the behavior data and the product test data; Feature extraction module: used to extract characteristic rules from the behavior data using a convolutional neural network to obtain characteristic data, and to perform behavioral semantic recognition on the characteristic data to obtain semantic data; including: The image features of the behavior data are extracted using a convolutional neural network to obtain the image features, and the image features are screened. The expression is: ， ， ， The first image feature is , the second image feature is , the binary XOR operation algorithm is , the total number of digits whose statistical operation result is 1 is , the Hamming distance between the first image feature and the second image feature is H, and the minimum value of the Hamming distance is , the remainder is %, and the maximum value of the Hamming distance is , the Hamming distance between the b-th image feature pair is , the automatic change parameters are , set the constants to , ; To satisfy the Hamming distance The image features of the judgment conditions are further screened, and the expression is: ， ， The window centered on the image feature is G, and the offset coordinates inside the window G are , the first image feature at the s+1th moment is , the first image feature at the sth moment is , the first image feature and The similarity score is U, and the first image feature of the behavior data at the s+1th moment is , the first image feature of the behavior data at the sth moment is , the similarity of the first image feature of the first behavior data and the second behavior data is ; Given a similarity threshold, when the similarity U is less than the similarity threshold, the image feature is output as the key feature; Calculate the degree of change of behavioral data: ， The bth key feature at the s+1th moment is , the bth key feature at the sth moment is , the number of key features is , the control coefficient is , the Euclidean norm is , adjust the parameters to , the similarity of the bth key feature between the sth moment and the s+1th moment is , the change degree of the bth behavior data between the sth moment and the s+1th moment is ; Classify the degree of change of behavioral data into groups and output them as characteristic data; Weighted construction module: used for associating the product test data with the semantic data to obtain associated data, performing directed weighting according to the associated data to obtain data weights, and constructing a satisfaction function using the semantic data and the data weights; Model building module: used to build a user satisfaction simulation model according to the satisfaction function, optimize the user satisfaction simulation model, input the data to be simulated into the user satisfaction simulation model, and output the simulation results.

Images