An interaction system and device for an e-commerce platform

Abstract

The present invention provides an interaction system and device for an e-commerce platform. The system performs the following steps: obtaining user behavior data and corresponding commodity information to generate a training sample set; classifying the user behavior data in the training sample set to obtain the type of each user behavior data; concentrating each user behavior data and the commodity information of each type in the behavior data to train a classification neural network model to obtain a user behavior classification model; correspondingly, feeding back and training a multi-channel neural network model to obtain a commodity recommendation model; inputting the to-be-interacted user behavior data into the user behavior classification model to obtain the type of the to-be-interacted user behavior; and inputting the to-be-interacted user behavior data into the commodity recommendation model according to the type of the to-be-interacted user behavior to obtain a commodity recommendation for the user behavior. The present invention can improve user satisfaction and purchase conversion rate and optimize the user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of e-commerce, and more specifically, the present invention relates to an interaction system and device for an e-commerce platform. Background Art

[0002] In the field of e-commerce, with the rapid development of Internet technology and the explosion of user numbers, how to effectively analyze user behavior and provide personalized product recommendations has become an important means for e-commerce platforms to improve user experience and increase sales. Existing technologies usually rely on simple algorithms or rules, such as making recommendations based on users' purchase history, browsing records, etc. These methods often fail to deeply understand users' complex preferences, resulting in low accuracy and personalization of recommendation results. In addition, the diversity and complexity of user behavior data also pose challenges to behavior analysis, and traditional processing methods often cannot effectively utilize all the information of these data.

[0003] In the process of implementing the embodiments of the present invention, the inventors found that there are at least the following problems or defects in the existing technologies: When existing e-commerce interaction systems process user behavior data, they often ignore the diversity and dynamic changes of user behavior, resulting in the inability to accurately capture users' real needs and preference changes. At the same time, most existing recommendation systems adopt a single recommendation model, lacking in-depth exploration of multi-dimensional features of user behavior types and product information, making the accuracy and personalization of recommendation results to be improved. In addition, when existing systems process large-scale user behavior data, they often face problems of computing efficiency and real-time performance, and it is difficult to meet the requirements of e-commerce platforms for high efficiency and real-time performance. Summary of the Invention

[0004] The present invention provides an interaction system and device for an e-commerce platform.

[0005] In the first aspect of the present invention, there is provided an interaction system for an e-commerce platform, including:

[0006] Obtain user behavior data and corresponding product information, and generate a training sample set based on the user behavior data and the corresponding product information;

[0007] Classify the user behavior data in the training sample set based on user preferences to obtain the type of each user behavior data;

[0008] Train a classification neural network model based on each user behavior data in the sample set and the product information of each type in the behavior data to obtain a user behavior classification model;

[0009] Train a multi-channel neural network model based on each user behavior data in the sample set, the product information of each type in the behavior data, and the feedback corresponding to the user behavior to obtain a product recommendation model;

[0010] Input the user behavior data to be interacted into the user behavior classification model to obtain the type of the user behavior to be interacted; according to the type of the user behavior to be interacted, input the user behavior data to be interacted into the commodity recommendation model to obtain the commodity recommendation of the user behavior.

[0011] Further, classify the user behavior data in the training sample set based on user preferences, including: for each user behavior data, take the first behavior with corresponding commodity feedback as the first category; traverse each behavior in turn, and calculate the preference change rate between the current behavior and the previous behavior based on the commodity information corresponding to the user behavior.

[0012] If the preference change rate is greater than the threshold, the current behavior and the previous behavior are of different types; otherwise, the current behavior and the previous behavior are of the same type.

[0013] Further, the preference change rate includes the price change rate and the category change rate ;

[0014] Calculating the preference change rate between the current behavior and the previous behavior based on the commodity information corresponding to the user behavior includes: calculating the statistical features of the preference corresponding to each commodity information of the user behavior.

[0015] Use the following formula to calculate the preference change rate between the current behavior and the previous behavior: price change rate

[0016]

[0017] category change rate

[0018]

[0019] where and respectively represent the commodity prices corresponding to the current behavior and the previous behavior, and respectively represent the commodity categories corresponding to the current behavior and the previous behavior.

[0020] Further, the number of input channels of the multi-channel neural network model is the same as the number of behavior types;

[0021] Training the multi-channel neural network model to obtain the commodity recommendation model based on each user behavior data in the sample set, the commodity information of each type in the behavior data, and the feedback corresponding to the user behavior includes: inputting the behaviors with the same type in each user behavior data into the same input channel to train the multi-channel neural network model.

[0022] Further, use the following formula to calculate the loss of the multi-channel neural network model:

[0023]

[0024] Among them, represents the commodity recommendation matrix corresponding to the th behavior type predicted by the model, represents the gold standard commodity recommendation matrix corresponding to the th behavior type, represents the number of behavior categories, and represent weight parameters, represents the regularization parameter, represents the total sample loss.

[0025] Furthermore, a training sample set is generated based on the user behavior data and the corresponding commodity information, including: resampling the user behavior data and the corresponding commodity information to standard voxels by using linear interpolation;

[0026] Normalizing the voxel values of the user behavior data;

[0027] Performing image segmentation on the normalized user behavior data to obtain a user behavior mask, calculating the minimum bounding cuboid of the user behavior mask, and extracting the user behavior data and the corresponding commodity information in the minimum bounding cuboid to generate a training sample set.

[0028] Furthermore, the multi-channel neural network model includes a multi-channel convolutional layer, a downsampling convolutional module, and an upsampling convolutional module connected in sequence;

[0029] The multi-channel convolutional layer is used to extract multi-channel feature images from multiple input channels by using convolution; the downsampling convolutional module is used to perform feature extraction at different levels on the multi-channel feature images, and the upsampling convolutional module is used to perform upsampling on the extracted features and output a recommendation result;

[0030] The downsampling convolutional module includes a plurality of downsampling convolutional units, and each downsampling convolutional unit includes a three-dimensional convolutional layer, a LeakyReLU layer, a batch normalization layer, and a max pooling layer connected in sequence;

[0031] The upsampling convolutional module includes a plurality of upsampling convolutional units, and each upsampling convolutional unit includes a three-dimensional convolutional layer, a LeakyReLU layer, a batch normalization layer, and a transposed convolutional layer connected in sequence.

[0032] In the second aspect of the present invention, an interaction device for an e-commerce platform is provided, including:

[0033] A sample set generation module, configured to obtain user behavior data and corresponding product information, and generate a training sample set based on the user behavior data and the corresponding product information;

[0034] A behavior classification module, configured to classify the user behavior data in the training sample set based on user preferences to obtain the type of each user behavior data;

[0035] A classification model training module, configured to train a classification neural network model based on the product information of each type in each user behavior data in the sample set to obtain a user behavior classification model;

[0036] A recommendation model training module, configured to train a multi-channel neural network model based on each user behavior data in the sample set, the product information of each type in the behavior data, and the feedback corresponding to the user behavior to obtain a product recommendation model;

[0037] An automatic interaction module, configured to input the user behavior data to be interacted into the user behavior classification model to obtain the type of the user behavior to be interacted; and input the user behavior data to be interacted into the product recommendation model according to the type of the user behavior to be interacted to obtain a product recommendation for the user behavior.

[0038] Further, the behavior classification module classifies the user behavior data in the training sample set by the following steps: for each user behavior data, taking the first behavior with corresponding product feedback as the first category; traversing each behavior in turn, and calculating the preference change rate between the current behavior and the previous behavior based on the product information corresponding to the user behavior;

[0039] If the preference change rate is greater than the threshold, the current behavior and the previous behavior are of different types; otherwise, the current behavior and the previous behavior are of the same type.

[0040] Further, the preference change rate includes a price change rate and a category change rate ;

[0041] Calculating the preference change rate between the current behavior and the previous behavior based on the product information corresponding to the user behavior includes: calculating the statistical features of the preference corresponding to each product information of the user behavior;

[0042] The following formula is used to calculate the preference change rate between the current behavior and the previous behavior: price change rate

[0043]

[0044] category change rate

[0045]

[0046] where, and respectively represent the commodity prices corresponding to the current row and the previous row, and respectively represent the commodity categories corresponding to the current row and the previous row.

[0047] The above embodiments of the present invention have at least the following beneficial effects: By adopting advanced machine learning technology, the e-commerce platform interaction system can deeply analyze and classify user behavior data, so as to more accurately identify the types of user behavior and preference changes. The system uses a multi-channel neural network model, which can comprehensively consider multiple dimensions of user behavior, including price, category, etc., and provide more accurate commodity recommendations for users. In addition, by updating user behavior data and commodity information in real time, the system can dynamically adjust the recommendation strategy to adapt to the changes in user preferences, thereby improving the relevance of recommendations and user satisfaction.

[0048] By this method, the system can not only improve the shopping experience of users, increase user stickiness, but also improve the conversion rate of commodities through accurate recommendations, bringing higher economic benefits to the e-commerce platform. At the same time, the automated interaction mechanism of the system can reduce manual intervention, lower operating costs, and improve the efficiency of processing large-scale data, enabling the e-commerce platform to better respond to market changes and the rapid growth of user needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] By referring to the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present invention will become readily understood. In the drawings, several embodiments of the present invention are shown by way of illustration and not limitation, wherein:

[0050] Figure 1 is a schematic flow chart of an e-commerce platform interaction system provided by an embodiment of the present invention;

[0051] Figure 2 is a schematic structural diagram of an e-commerce platform interaction device provided by an embodiment of the present invention;

[0052] Figure 3 schematically shows a schematic structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] The principles and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are given only to enable those skilled in the art to better understand and then implement the present invention, and do not limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to be able to fully convey the scope of the present invention to those skilled in the art.

[0054] Those skilled in the art know that the embodiments of the present invention can be implemented as a system, a device, an equipment, a method or a computer program product. Therefore, the present invention can be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0055] It should be noted that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction and does not have any limiting meaning.

[0056] The following refers to Figure 1 , Figure 1 which is a schematic flowchart of an e-commerce platform interaction system provided for an embodiment of the present invention. As Figure 1 shown, an interaction system 100 for an e-commerce platform includes:

[0057] Step 101, obtaining user behavior data and corresponding product information, and generating a training sample set based on the user behavior data and the corresponding product information;

[0058] Step 102, classifying the user behavior data in the training sample set based on user preferences to obtain the type of each user behavior data;

[0059] Step 103, training a classification neural network model based on each user behavior data in the sample set and the product information of each type in the behavior data to obtain a user behavior classification model;

[0060] Step 104, training a multi-channel neural network model based on each user behavior data in the sample set, the product information of each type in the behavior data, and the feedback corresponding to the user behavior to obtain a product recommendation model;

[0061] Step 105, inputting the to-be-interacted user behavior data into the user behavior classification model to obtain the type of the to-be-interacted user behavior; and inputting the to-be-interacted user behavior data into the product recommendation model according to the type of the to-be-interacted user behavior to obtain a product recommendation for the user behavior.

[0062] It should be noted that the system first obtains user behavior data and corresponding product information, and these data include the user's browsing history, purchase records, click behaviors, etc., as well as the product details related to these behaviors, such as price, category, description, etc.

[0063] Specifically, the user behavior data can be collected through log files, database queries, etc., while the product information comes from the product database of the e-commerce platform. In the process of generating the training sample set, the system will adopt certain data preprocessing technologies, such as data cleaning, denoising, etc., to ensure the quality and usability of the data. In addition, the system will adopt some feature engineering technologies, such as feature selection, feature extraction, etc., to enhance the performance of the model.

[0064] Preferably, when processing user behavior data, the system will adopt some advanced data processing technologies, such as sequence models in deep learning, to capture the temporal characteristics of user behavior. At the same time, the system will introduce some regularization technologies, such as L1 and L2 regularization, to prevent the model from overfitting and improve the generalization ability of the model. In addition, the system will also adopt some optimization algorithms, such as Stochastic Gradient Descent (SGD), Adam, etc., to accelerate the training process of the model.

[0065] It should be noted that when classifying user behavior data, the system described in this embodiment adopts a method based on user preferences. This includes classifying user behavior data according to certain rules to identify the user's behavior patterns and preference changes.

[0066] Specifically, the system will define a threshold to determine whether the preference change between user behaviors is significant. If the preference change rate between two consecutive behaviors exceeds this threshold, they are considered to belong to different types. The calculation of the preference change rate will involve multiple dimensions such as the price change rate and the category change rate, and these dimensions can be quantified through statistical methods.

[0067] More specifically, when calculating the preference change rate, the system will adopt some advanced mathematical formulas and algorithms. For example, the price change rate will be calculated using the relative change in price, and the category change rate will consider the similarity of product categories. In addition, the system will also introduce some machine learning technologies, such as clustering analysis, to automatically determine the division of behavior types.

[0068] It should be noted that the multi-channel neural network model described in this embodiment is the core component for the system to implement product recommendations. This model can process data from multiple input channels and output corresponding recommendation results.

[0069] Specifically, the number of input channels of the multi-channel neural network model will match the number of types of user behavior. During the training process, the system will input user behavior data of the same type into the corresponding input channels to train the model to identify and recommend products of the corresponding type.

[0070] More specifically, the system adopts some advanced neural network structures, such as downsampling convolutional modules and upsampling convolutional modules, to extract and fuse multi-channel features. These structures can help the model better understand the complex patterns in user behavior data and improve the accuracy of recommendation results. In addition, the system will also adopt some regularization technologies, such as Dropout, to prevent the model from overfitting.

[0071] In some embodiments, classifying user behavior data in a training sample set based on user preferences includes: for each piece of user behavior data, taking the first behavior with corresponding product feedback as the first category; traversing each behavior in sequence, and calculating the preference change rate between the current behavior and the previous behavior based on the product information corresponding to the user behavior;

[0072] If the preference change rate is greater than the threshold, the current behavior and the previous behavior are of different types; otherwise, the current behavior and the previous behavior are of the same type.

[0073] It should be noted that when classifying user behavior data, the e-commerce platform interaction system in this embodiment pays special attention to the change of user preferences. The system calculates the preference change rate by analyzing the product information in the user behavior data, so as to classify user behavior into different types.

[0074] Specifically, the preference change rate of users includes the price change rate and the category change rate. The price change rate refers to the degree of difference in the prices of the products concerned by the user in two consecutive behaviors; the category change rate refers to the degree of change in the product categories concerned by the user. These two indicators can help the system understand the dynamic changes of user preferences, so as to classify user behavior more accurately.

[0075] Preferably, when calculating the price change rate and the category change rate, the system will use specific mathematical formulas. For example, the price change rate can be obtained by calculating the relative change in the product price in two consecutive behaviors, while the category change rate is calculated by comparing the similarity of product categories. In addition, the system will set a threshold. When the change rate exceeds this threshold, it is considered that the user's behavior type has changed.

[0076] It should be noted that the system in this embodiment adopts a threshold-based classification method when processing user behavior data. This method judges whether user behaviors belong to the same type by setting a threshold.

[0077] Specifically, the system will set this threshold according to historical data and the statistical characteristics of user behavior. For example, the system will calculate the historical average of the price change rate and the category change rate in user behavior, and then use this average as a reference for the threshold.

[0078] More specifically, the system will use some machine learning techniques to dynamically adjust this threshold. For example, the system will adopt an online learning method to update the threshold in real time according to the latest user behavior data to adapt to the change of user preferences.

[0079] It should be noted that when classifying user behaviors in this embodiment, the system particularly considers the relationship between user behaviors and commodity information. The system calculates the change rate of user preferences by analyzing the commodity information corresponding to user behaviors.

[0080] Specifically, the system will collect detailed information in user behavior data, such as the commodity information browsed, clicked, and purchased by the user, and then calculate the price change rate and category change rate based on this information.

[0081] Preferably, when calculating these change rates, the system will adopt some advanced data processing technologies. For example, the system will use time series analysis to process the price change rate and clustering analysis to process the category change rate. In addition, the system will also adopt some statistical methods, such as variance analysis, to evaluate the significance of the change rate.

[0082] In some embodiments, the preference change rate includes the price change rate and the category change rate ;

[0083] Calculating the preference change rate between the current behavior and the previous behavior based on the commodity information corresponding to the user behavior includes: calculating the statistical characteristics of the preference corresponding to each commodity information of the user behavior;

[0084] The following formula is used to calculate the preference change rate between the current behavior and the previous behavior: price change rate

[0085]

[0086] category change rate

[0087]

[0088] where and respectively represent the commodity prices corresponding to the current behavior and the previous behavior, and respectively represent the commodity categories corresponding to the current behavior and the previous behavior.

[0089] It should be noted that when calculating the user preference change rate, the e-commerce platform interaction system in this embodiment particularly focuses on the changes in two dimensions: price and category. By accurately calculating the change rates of these two dimensions, the system can more accurately capture the subtle changes in user preferences.

[0090] Specifically, the price change rate is obtained by calculating the relative change in the commodity prices concerned by the user in two consecutive behaviors. The category change rate is calculated by comparing the category differences of the commodities concerned by the user in two consecutive behaviors. The calculation of these two change rates can help the system more carefully understand the user's behavior patterns and preference changes.

[0091] Preferably, when calculating the price change rate and the category change rate, the system will adopt specific mathematical formulas. The calculation formula for the price change rate is:

[0092]

[0093] The calculation formula for the category change rate is:

[0094]

[0095] Among them, and respectively represent the commodity prices corresponding to the current behavior and the previous behavior, and respectively represent the commodity categories corresponding to the current behavior and the previous behavior.

[0096] Specifically, the system will collect detailed information in the user behavior data, such as the commodity information browsed, clicked, and purchased by the user, and then calculate the price change rate and the category change rate based on this information. These statistical features include the price range, category distribution, etc. of the commodities.

[0097] Furthermore, when calculating these change rates, the system will adopt some advanced data processing technologies. For example, the system will adopt time series analysis to process the price change rate and clustering analysis to process the category change rate. In addition, the system will also adopt some statistical methods, such as variance analysis, to evaluate the significance of the change rate.

[0098] More specifically, the system will dynamically adjust the calculation formulas of the price change rate and the category change rate according to the user's real-time behavior data. For example, the system will adjust the weight of the change rate according to the frequency and intensity of the user's behavior to better reflect the change of the user's preference.

[0099] Even further, the system will adopt some machine learning technologies to achieve this dynamic adjustment. For example, the system will adopt a reinforcement learning algorithm to optimize the calculation formula of the change rate, so that the system can automatically adjust the calculation strategy according to the user's feedback, thereby more accurately capturing the change of the user's preference.

[0100] In some embodiments, the number of input channels of the multi-channel neural network model is the same as the number of behavior types;

[0101] Training a multi-channel neural network model based on each user behavior data, the commodity information of each type in the behavior data, and the feedback corresponding to the user behavior to obtain a commodity recommendation model, including: inputting behaviors of the same type in each user behavior data into the same input channel to train the multi-channel neural network model.

[0102] It should be noted that the e-commerce platform interaction system in this embodiment uses a multi-channel neural network model to process user behavior data, and the number of input channels of this model matches the number of types of user behavior. Such a design enables the system to conduct in-depth analysis and learning for different types of user behavior respectively.

[0103] Specifically, the multi-channel neural network model is a deep learning model that can process data from multiple input sources. In this system, each input channel is dedicated to processing a specific type of user behavior data. For example, if the user behavior data includes three types: browsing, clicking, and purchasing, then the model will have three corresponding input channels.

[0104] Preferably, when designing the multi-channel neural network model, the system will adopt different network structures to adapt to different types of user behavior data. For example, for behavior data containing time series information, recurrent neural network structures such as LSTM or GRU will be adopted; while for behavior data such as images or texts, convolutional neural network CNN or Transformer structures will be adopted.

[0105] Specifically, the system will adopt a special data preprocessing method to group the user behavior data by type and then input them into the corresponding channels of the model respectively. For example, all click behavior data will be input into the first channel, all browsing behavior data will be input into the second channel, and so on.

[0106] Preferably, the system will adopt some advanced data processing techniques to enhance the performance of the model. For example, data augmentation techniques will be used to generate more training samples, or regularization techniques will be used to prevent the model from overfitting. In addition, the system will also adopt some optimization algorithms, such as Adam or RMSprop, to accelerate the training process of the model.

[0107] It should be noted that when training the multi-channel neural network model, the system in this embodiment will consider the complexity of user behavior data and commodity information. The system will use this information to train the model so that the model can better understand and predict user behavior.

[0108] Specifically, the system will adopt a complex network structure, such as a deep convolutional neural network, to process high-dimensional user behavior data and commodity information. This network structure can automatically extract key features in the data and learn the complex relationships between the data.

[0109] Preferably, the system will adopt some advanced feature fusion techniques to integrate information from different channels. For example, the attention mechanism will be used to dynamically adjust the importance of different channels, or the gating mechanism will be used to control the flow of information. These techniques can help the model learn and predict user behavior more effectively.

[0110] In some embodiments, the following formula is used to calculate the loss of the multi-channel neural network model:

[0111]

[0112] Wherein, represents the commodity recommendation matrix corresponding to the th behavior type predicted by the model, represents the gold standard commodity recommendation matrix corresponding to the th behavior type, represents the number of behavior categories, and represent weight parameters, represents the regularization parameter, represents the total loss of the samples.

[0113] Specifically, the calculation of the loss function involves the difference between the commodity recommendation matrix predicted by the model and the gold standard commodity recommendation matrix. This difference is quantified by calculating the error between the predicted matrix and the true matrix. The smaller the error, the better the prediction performance of the model.

[0114] Preferably, the system will adopt some advanced error measurement methods, such as mean square error MSE or cross-entropy loss, to calculate the error between the predicted matrix and the true matrix. In addition, the system will also introduce some regularization terms, such as L1 or L2 regularization, to prevent the model from overfitting.

[0115] It should be noted that the loss function in this embodiment includes multiple components, and each component targets different prediction behavior types of the model. Such a design enables the system to optimize the prediction performance of the model for different types of user behaviors respectively.

[0116] Specifically, the loss function will contain multiple terms, and each term corresponds to a behavior type. For example, if the system needs to predict three different types of user behaviors, then the loss function will have three corresponding terms. Each term will calculate the error of this type of prediction and incorporate it into the calculation of the total loss.

[0117] Preferably, the system will adopt some weight parameters to balance the importance of different behavior types. For example, if the prediction of certain behavior types is more critical to the final recommendation result, then the loss terms of these behavior types will be assigned higher weights.

[0118] It should be noted that the loss function in this embodiment also considers the regularization term to improve the generalization ability of the model. The regularization term can help the model avoid overfitting, so that it can also maintain good prediction performance on new and unseen data.

[0119] More specifically, the regularization term will include penalties on the model parameters, such as the L1 norm or L2 norm of the parameters. These penalty terms can limit the scale of the model parameters, thus preventing the model from becoming overly complex.

[0120] Furthermore, the system will adopt some advanced regularization techniques, such as Dropout or Batch Normalization, to further improve the generalization ability of the model. These techniques can dynamically adjust the complexity of the model during training, thus avoiding overfitting while improving the model performance.

[0121] In some embodiments, generating a training sample set based on the user behavior data and the corresponding commodity information includes: resampling the user behavior data and the corresponding commodity information to a standard voxel using linear interpolation;

[0122] Normalizing the voxel values of the user behavior data;

[0123] Performing image segmentation on the normalized user behavior data to obtain a user behavior mask, calculating the minimum bounding box of the user behavior mask, and extracting the user behavior data and the corresponding commodity information in the minimum bounding box to generate a training sample set.

[0124] Specifically, resampling adjusts the user behavior data and the corresponding commodity information to a unified standard voxel size through linear interpolation. Normalization adjusts the numerical range of these data to a specific interval, such as 0 to 1, to eliminate the numerical differences between different features. Image segmentation segments the voxel values in the user behavior data to identify the key regions of the user behavior.

[0125] Preferably, the system will adopt some advanced image processing techniques to improve the accuracy of data preprocessing. For example, more complex interpolation methods, such as cubic spline interpolation, are used to improve the quality of resampling. At the same time, different methods, such as min-max normalization or Z-score standardization, are adopted in the normalization process to adapt to different types of data distributions.

[0126] It should be noted that when generating the training sample set in this embodiment, the system pays special attention to the voxel values of the user behavior data. By normalizing and performing image segmentation on the voxel values, the system can more accurately identify the key features of the user behavior.

[0127] Specifically, the normalization of the voxel values can be performed according to the following formula:

[0128]

[0129] Image segmentation identifies different regions in the voxel values and calculates the minimum bounding box of the user behavior mask.

[0130] Preferably, the system will adopt some advanced image segmentation algorithms, such as threshold - based segmentation or region - based segmentation, to improve the accuracy of segmentation. In addition, the system will also adopt some morphological operations, such as erosion and dilation, to optimize the quality of the user behavior mask.

[0131] It should be noted that when the system in this embodiment generates the training sample set, it will also extract the user behavior data and the corresponding product information in the minimum bounding box of the user behavior mask. This step helps the system to more accurately locate the key areas of user behavior and use them for model training.

[0132] Specifically, the minimum bounding box can be obtained by calculating the bounding box of the user behavior mask. Extracting the user behavior data and product information within this area can ensure that the data in the training sample set is the most representative and informative.

[0133] Preferably, the system will adopt some advanced geometric processing techniques to optimize the calculation of the minimum bounding box. For example, using the convex hull algorithm to determine the boundary of the user behavior mask, or using more complex geometric algorithms to improve the accuracy of the bounding box. In addition, the system will also introduce some machine learning techniques, such as clustering analysis, to automatically identify and extract the key user behavior areas.

[0134] In some embodiments, the multi - channel neural network model includes a multi - channel convolutional layer, a down - sampling convolutional module, and an up - sampling convolutional module connected in sequence;

[0135] The multi - channel convolutional layer is used to extract multi - channel feature images from multiple input channels by convolution; the down - sampling convolutional module is used to perform feature extraction at different levels on the multi - channel feature images, and the up - sampling convolutional module is used to upsample the extracted features and output the recommendation result;

[0136] The down - sampling convolutional module includes a plurality of down - sampling convolutional units, and each down - sampling convolutional unit includes a three - dimensional convolutional layer, a LeakyReLU layer, a batch normalization layer, and a max - pooling layer connected in sequence;

[0137] The up - sampling convolutional module includes a plurality of up - sampling convolutional units, and each up - sampling convolutional unit includes a three - dimensional convolutional layer, a LeakyReLU layer, a batch normalization layer, and a transposed convolutional layer connected in sequence.

[0138] Specifically, the multi - channel convolutional layer can process data with multiple input channels, and each channel corresponds to data of a type of user behavior. The down - sampling convolutional module is used to extract features at different levels, and the up - sampling convolutional module is used to upsample the extracted features in order to output the final recommendation result.

[0139] Preferably, the system will use some advanced convolutional neural network technology to optimize the performance of the model. For example, deep separable convolution is used to reduce the number of model parameters and computational complexity. At the same time, the system will also introduce an attention mechanism to enhance the model's ability to recognize key features.

[0140] It should be noted that the multi-channel convolution layer in this embodiment is used to extract multi-channel feature images from multiple input channels. These feature images contain important information of user behavior data and provide a basis for subsequent recommendations.

[0141] Specifically, the multi-channel convolution layer uses convolution kernels of different sizes to capture features of different scales. In addition, the system sets different numbers of channels to adapt to different types of user behavior data.

[0142] Furthermore, the system will use some regularization techniques, such as Dropout or Batch Normalization, to improve the generalization ability of the model. At the same time, the system will also use some optimization algorithms, such as Adam or RMSprop, to accelerate the training process of the model.

[0143] It should be noted that the downsampling convolution module in this embodiment includes multiple downsampling convolution units, each of which is composed of a three-dimensional convolution layer, a LeakyReLU layer, a batch normalization layer, and a maximum pooling layer connected in sequence. Such a design helps the model extract deep features from multi-channel feature images.

[0144] Specifically, the 3D convolutional layer is able to process data with three dimensions, the LeakyReLU layer is used to introduce nonlinearity, the batch normalization layer helps to speed up the training process, and the maximum pooling layer is used to reduce the spatial dimension of the features.

[0145] More specifically, the system will use some advanced pooling techniques, such as adaptive pooling, to better capture the dynamic range of features. At the same time, the system will also introduce some advanced activation functions, such as ELU or SELU, to improve the nonlinear expression ability of the model.

[0146] The above-mentioned embodiments of the present invention have the following beneficial effects: the e-commerce platform interaction system of the present invention can effectively integrate user behavior data and product information, generate training sample sets, and realize personalized product recommendations through user behavior classification models and product recommendation models. The system can automatically learn and adjust recommendation strategies based on user behavior data and feedback, thereby providing users with more accurate and personalized product recommendations. In addition, by adopting classification neural network models and multi-channel neural network models, the system can more deeply mine the potential information in user behavior data and improve the performance and accuracy of the recommendation system.

[0147] The e-commerce platform interaction system can enhance the user interaction experience, improve user satisfaction and loyalty. At the same time, the automation and intelligence features of the system can reduce the operating costs of the e-commerce platform and improve the operating efficiency. The system can also quickly respond to market changes and user needs by analyzing and processing user behavior data in real time, bringing higher commercial value and competitive advantages to the e-commerce platform.

[0148] As Figure 2 shown, an interaction device 200 for an e-commerce platform in some embodiments, the device 200 includes:

[0149] A sample set generation module 201, configured to obtain user behavior data and corresponding product information, and generate a training sample set based on the user behavior data and the corresponding product information;

[0150] A behavior classification module 202, configured to classify the user behavior data in the training sample set based on user preferences to obtain the type of each user behavior data;

[0151] A classification model training module 203, configured to train a classification neural network model based on the product information of each type in each user behavior data in the sample set to obtain a user behavior classification model;

[0152] A recommendation model training module 204, configured to train a multi-channel neural network model based on each user behavior data in the sample set, the product information of each type in the behavior data, and the feedback corresponding to the user behavior to obtain a product recommendation model;

[0153] An automatic interaction module 205, configured to input the user behavior data to be interacted into the user behavior classification model to obtain the type of the user behavior to be interacted; and input the user behavior data to be interacted into the product recommendation model according to the type of the user behavior to be interacted to obtain a product recommendation for the user behavior.

[0154] It can be understood that the modules described in the e-commerce platform interaction device 200 correspond to the respective steps in the e-commerce platform interaction system described with reference to Figure 1 description. Therefore, the operations, features, and beneficial effects described above for the e-commerce platform interaction system also apply to the e-commerce platform interaction device 200 and the modules included therein, and will not be elaborated here.

[0155] In some embodiments, the behavior classification module classifies the user behavior data in the training sample set by the following steps: for each user behavior data, the first behavior with corresponding product feedback is used as the first category; each behavior is traversed in sequence, and the preference change rate between the current behavior and the previous behavior is calculated based on the product information corresponding to the user behavior;

[0156] If the preference change rate is greater than the threshold, the current behavior and the previous behavior are of different types; otherwise, the current behavior and the previous behavior are of the same type.

[0157] It should be noted that the behavior classification module of the e-commerce platform interaction device in this embodiment adopts a series of steps to classify the user behavior data in the training sample set. This process involves grouping the user behavior data according to its relevance to the product feedback to identify different user behavior types.

[0158] Specifically, the behavior classification module first marks the first behavior with corresponding product feedback in each user behavior data as the first category. Then, the module will sequentially traverse each behavior and calculate the preference change rate between the current behavior and the previous behavior based on the product information corresponding to the user behavior.

[0159] Preferably, the system will adopt some advanced data analysis techniques to improve the accuracy of behavior classification. For example, clustering analysis is used to automatically identify patterns in the behavior data, or time series analysis is used to consider the time dependence of the behavior data.

[0160] It should be noted that when calculating the preference change rate, the behavior classification module in this embodiment will consider the price change rate and the category change rate. These change rates are calculated based on the product information corresponding to the user behavior and are used to measure the dynamic change of user preferences.

[0161] Specifically, the calculation of the price change rate and the category change rate involves complex mathematical formulas. For example, the price change rate is calculated by comparing the product prices corresponding to the current behavior and the previous behavior, while the category change rate is obtained by comparing the category identifiers of the products.

[0162] Preferably, the system will introduce some machine learning algorithms to optimize the calculation process of the change rate. For example, support vector machine (SVM) is used to determine the thresholds for price and category changes, or neural networks are used to learn the relationship between the change rate and the user behavior type.

[0163] It should be noted that when determining the user behavior type, the behavior classification module in this embodiment will set a threshold to judge whether the preference change rate is significant. If the change rate exceeds this threshold, it is considered that the current behavior and the previous behavior belong to different types; otherwise, they belong to the same type.

[0164] Specifically, the setting of the threshold will be based on the statistical analysis of historical data. The system will calculate the median or average of the historical preference change rates and use it as the initial setting of the threshold.

[0165] Furthermore, the system will adopt some adaptive methods to dynamically adjust the threshold. For example, it adjusts the threshold according to the real-time changes in user behavior data, or uses online learning algorithms to update the threshold in real time to adapt to the dynamic changes in user behavior. In addition, the system will also introduce some anomaly detection mechanisms to handle behavior data that does not conform to the normal pattern.

[0166] In some embodiments, the preference change rate includes a price change rate and a category change rate ;

[0167] Calculating the preference change rate between the current behavior and the previous behavior based on the commodity information corresponding to the user behavior includes: calculating the statistical features of the preference corresponding to each commodity information of the user behavior;

[0168] The following formula is used to calculate the preference change rate between the current behavior and the previous behavior: price change rate

[0169]

[0170] category change rate

[0171]

[0172] where and respectively represent the commodity prices corresponding to the current behavior and the previous behavior, and respectively represent the commodity categories corresponding to the current behavior and the previous behavior.

[0173] It should be noted that when calculating the preference change rate of the user behavior data, the behavior classification module of the e-commerce platform interaction device in this embodiment pays special attention to these two key indicators: the price change rate and the category change rate. These two indicators help the module more accurately identify the type of user behavior, so as to provide more accurate user preference information for subsequent commodity recommendations.

[0174] Specifically, the price change rate refers to the degree of difference in the prices of the commodities concerned by the user in two consecutive behaviors. The category change rate refers to the degree of difference in the categories of the commodities concerned by the user. The calculation of these two indicators is based on the commodity information corresponding to the user behavior, such as attributes like price and category.

[0175] Preferably, the system will adopt some advanced mathematical models and algorithms to calculate these two change rates. For example, statistical methods are used to analyze the historical change trends of prices and categories, or machine learning algorithms are used to predict the sensitivity of users to price and category changes.

[0176] It should be noted that the calculation of the preference change rate in this embodiment includes the calculation of the statistical characteristics of the preferences corresponding to each commodity information of the user's behavior. This means that the system not only considers the changes in price and category, but also considers other commodity attributes, such as brand, evaluation, etc., in order to obtain more comprehensive user preference information.

[0177] Specifically, the system will use feature extraction techniques to identify and quantify the user's preferences for different commodity attributes. For example, natural language processing techniques are used to analyze the user's feedback on commodity descriptions, or image recognition techniques are used to analyze the user's preferences for commodity pictures.

[0178] Preferably, the system will use some advanced data analysis techniques to improve the accuracy of the preference change rate calculation. For example, deep learning algorithms are used to automatically learn the complex patterns of user preferences, or reinforcement learning algorithms are used to optimize the prediction model of user preferences.

[0179] It should be noted that when calculating the preference change rate, the system in this embodiment uses specific formulas to quantify the price change rate and the category change rate. These formulas provide a standardized method to measure the change in user preferences, thus helping the system to more accurately distinguish different types of user behavior.

[0180] Furthermore, the calculation formulas for the price change rate and the category change rate involve complex mathematical operations, such as ratio calculation, normalization, etc. The design of these formulas aims to ensure that the calculation of the change rate is both accurate and comparable.

[0181] Even further, the system will use some advanced mathematical optimization techniques to improve these formulas. For example, numerical analysis methods are used to optimize the parameters in the formulas, or machine learning techniques are used to automatically adjust the formulas to adapt to different user behavior patterns. In addition, the system will also introduce some regularization terms to prevent overfitting problems in the calculation process.

[0182] The following refers to Figure 3 , which shows a schematic structural diagram of the structure 300 of an electronic device suitable for implementing some embodiments of the present invention. The electronic devices in some embodiments of the present invention may include, but are not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Tablet Computers), PMPs (Portable Multimedia Players), in-vehicle terminals (such as in-vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The terminal devices shown are merely examples and should not impose any limitations on the functions and usage scopes of the embodiments of the present invention.

[0183] As Figure 3As shown, the electronic device 300 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 301, which may perform various appropriate actions and processes according to a program stored in the read-only memory (ROM) 302 or a program loaded from the storage device 308 into the random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304. The input / output (I / O) interface 305 is also connected to the bus 304.

[0184] Generally, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 may allow the electronic device 300 to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 3 the electronic device 300 with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. Instead, more or fewer devices may be implemented or had. Figure 3 Each block shown in may represent a device or, as needed, multiple devices.

[0185] Furthermore, the storage medium of the embodiments of the present application stores program instructions capable of implementing all the above methods. Among them, the program instructions may be stored in the above storage medium in the form of a software product, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods described in the various embodiments of the present application. And the foregoing storage medium includes: various media that can store program codes such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc, or a terminal device such as a computer, a server, a mobile phone, a tablet, etc.

[0186] The above description is only some preferred embodiments of the present invention and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by the specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by mutually replacing the above features with the technical features (but not limited to) having similar functions disclosed in the embodiments of the present invention.

Claims

1. An interactive system for an e-commerce platform, characterized in that: The system performs the following steps: Acquire user behavior data and corresponding product information, and generate a training sample set based on the user behavior data and the corresponding product information; Classify the user behavior data in the training sample set based on user preferences to obtain the type of each user behavior data; Based on each user behavior data in the sample set and each type of product information in the behavior data, a classification neural network model is trained to obtain a user behavior classification model; The product recommendation model is obtained by training a multi-channel neural network model based on the behavior data of each user in the sample set, the product information of each type in the behavior data, and the feedback corresponding to the user behavior; Inputting the behavior data of the user to be interacted into the user behavior classification model to obtain the type of the user behavior to be interacted; According to the type of the user behavior to be interacted, the user behavior data to be interacted is input into the product recommendation model to obtain product recommendations for the user behavior; Classifying the user behavior data in the training sample set based on user preferences, including: for each user behavior data, taking the first behavior with corresponding product feedback as the first category; Traverse each behavior in turn, and calculate the preference change rate between the current behavior and the previous behavior based on the product information corresponding to the user behavior; If the preference change rate is greater than the threshold, the current behavior and the previous behavior are of different types; otherwise, the current behavior and the previous behavior are of the same type.

2. The interactive system for an e-commerce platform according to claim 1, characterized in that: The preference change rate includes the price change rate and category change rate ; Calculating the preference change rate between the current behavior and the previous behavior based on the product information corresponding to the user behavior, including: calculating the statistical characteristics of the preference corresponding to each product information of the user behavior; The rate of change in preference between the current action and the previous action is calculated using the following formula: Price Change Rate Category change rate in, and Respectively represent the commodity prices corresponding to the current behavior and the previous behavior, and Respectively represent the product categories corresponding to the current behavior and the previous behavior.

3. The interactive system for an e-commerce platform according to claim 1, characterized in that: The number of input channels of the multi-channel neural network model is the same as the number of behavior types; A product recommendation model is obtained by training a multi-channel neural network model based on each user behavior data in a sample set, each type of product information in the behavior data, and feedback corresponding to the user behavior, including: inputting behaviors of the same type in each user behavior data into the same input channel to train the multi-channel neural network model.

4. The interactive system for an e-commerce platform according to claim 1, characterized in that: The loss of the multi-channel neural network model is calculated using the following formula: in, The model predicts the The product recommendation matrix corresponding to each behavior type, Indicates The gold standard product recommendation matrix corresponding to each behavior type, represents the number of behavior categories, and represents the weight parameter, represents the regularization parameter, Represents the total sample loss.

5. The interactive system for an e-commerce platform according to claim 1, characterized in that: Generating a training sample set based on the user behavior data and the corresponding product information includes: resampling the user behavior data and the corresponding product information to standard voxels using a linear interpolation method; Normalize the voxel values ​​of user behavior data; The normalized user behavior data is segmented to obtain a user behavior mask, the minimum circumscribed cuboid of the user behavior mask is calculated, the user behavior data and the corresponding product information in the minimum circumscribed cuboid are extracted, and a training sample set is generated.

6. The interactive system for an e-commerce platform according to claim 1, characterized in that: The multi-channel neural network model includes a multi-channel convolution layer, a downsampling convolution module and an upsampling convolution module connected in sequence; The multi-channel convolution layer is used to extract a multi-channel feature image from multiple input channels by using convolution; the downsampling convolution module is used to extract features of different levels from the multi-channel feature image, and the upsampling convolution module is used to upsample the extracted features and output the recommendation results; The downsampling convolution module includes a plurality of downsampling convolution units, each of which includes a three-dimensional convolution layer, a LeakyReLU layer, a batch normalization layer, and a maximum pooling layer connected in sequence; The upsampling convolution module includes multiple upsampling convolution units, each of which includes a three-dimensional convolution layer, a LeakyReLU layer, a batch normalization layer and a transposed convolution layer connected in sequence.

7. An interactive device for an e-commerce platform, applied to an interactive system for an e-commerce platform as claimed in any one of claims 1 to 6, characterized in that: Includes the following modules: A sample set generation module, used to obtain user behavior data and corresponding commodity information, and generate a training sample set based on the user behavior data and the corresponding commodity information; A behavior classification module is used to classify the user behavior data in the training sample set based on user preferences to obtain the type of each user behavior data; A classification model training module is used to train a classification neural network model based on each user behavior data in the sample set and each type of commodity information in the behavior data to obtain a user behavior classification model; A recommendation model training module is used to train a multi-channel neural network model based on each user behavior data in the sample set, each type of product information in the behavior data, and feedback corresponding to the user behavior to obtain a product recommendation model; An automatic interaction module, used for inputting the behavior data of the user to be interacted into the user behavior classification model to obtain the type of the user behavior to be interacted; According to the type of the user behavior to be interacted, the user behavior data to be interacted is input into the product recommendation model to obtain product recommendations for the user behavior.

8. The interactive device for an e-commerce platform according to claim 7, characterized in that: The behavior classification module classifies the user behavior data in the training sample set by using the following steps: for each user behavior data, the first behavior with corresponding product feedback is taken as the first category; each behavior is traversed in turn, and the preference change rate between the current behavior and the previous behavior is calculated based on the product information corresponding to the user behavior; If the preference change rate is greater than the threshold, the current behavior and the previous behavior are of different types; otherwise, the current behavior and the previous behavior are of the same type.

9. The interactive device for an e-commerce platform according to claim 8, characterized in that: The preference change rate includes the price change rate and category change rate ; Calculating the preference change rate between the current behavior and the previous behavior based on the product information corresponding to the user behavior, including: calculating the statistical characteristics of the preference corresponding to each product information of the user behavior; The rate of change in preference between the current action and the previous action is calculated using the following formula: Price Change Rate Category change rate in, and Respectively represent the commodity prices corresponding to the current behavior and the previous behavior, and Respectively represent the product categories corresponding to the current behavior and the previous behavior.

Images