Data transmission mode switching method, system, equipment and medium

Abstract

The invention relates to the technical field of transmission mode switching, and discloses a data transmission mode switching method, system and device and a medium, and the method comprises the steps: obtaining historical behavior data of a target user, carrying out the data standardization processing of the historical behavior data, and obtaining standard user behavior data; according to the standard user behavior data, extracting behavior characteristics fed back by the target user for each data transmission mode, and according to the behavior characteristics, constructing multi-dimensional behavior characteristics; calculating the data transmission reaching probability of each data transmission mode according to the multi-dimensional behavior characteristics; generating scene reaching modes under different data transmission scenes according to the data transmission reaching probabilities; and generating a data transmission mode switching strategy corresponding to the target user by using the scene reaching mode. According to the invention, the optimal data transmission mode can be judged by using the data transmission mode switching strategy, the dynamic decision of data pushing is realized, and the flexibility of the data transmission mode is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of transmission mode switching technology, and in particular to a data transmission mode switching method, system, device and medium. Background Technology

[0002] In internet marketing, reaching users is a crucial step, enabling timely information delivery and driving conversions. Therefore, improving user data reception is vital for business growth. Companies typically utilize various data transmission methods, such as phone calls, SMS, WeChat notifications, and app push notifications. Each method has its advantages. For instance, phone calls allow for real-time interaction, SMS and app push notifications deliver information quickly and passively, while WeChat notifications facilitate building long-term customer relationships and increasing user engagement. Current automated tools can achieve precise user segmentation and personalized push notifications based on user behavior, transmitting information through multiple channels.

[0003] In traditional marketing, businesses often rely on a single transmission method to reach customers. However, this reliance on a single method leads to inefficient information delivery. For example, phone calls may go unanswered, and text messages may be ignored, resulting in low efficiency. Furthermore, repetitive transmission through a single method can annoy users and reduce their brand satisfaction. At the same time, traditional data transmission methods struggle to automatically adjust to changes in user behavior. For instance, if a user deletes a WeChat account as a friend, it's impossible to immediately switch to another method, resulting in poor diversity and flexibility in current data transmission methods. Summary of the Invention

[0004] This invention provides a method, system, computer device, and medium for switching data transmission modes in artificial intelligence, in order to solve the problem of poor diversity and flexibility of existing data transmission modes.

[0005] Firstly, a method for switching data transmission modes is provided, including: Obtain historical behavior data of the target user, and perform data standardization processing on the historical behavior data to obtain standard user behavior data; Based on the standard user behavior data, extract the behavioral features of the target user's feedback to each data transmission method, and construct a multidimensional behavioral feature of the standard user behavior data based on the behavioral features. The data transmission reach probability of each of the data transmission methods is calculated based on the multidimensional behavioral characteristics. Based on the data transmission reach probability, the scenario reach method for the target user under different data transmission scenarios is generated; The data transmission method switching strategy corresponding to the target user is generated using the scenario-based outreach method.

[0006] Secondly, a data transmission mode switching system is provided, including: The data processing module is used to acquire the historical behavior data of the target user, and to perform data standardization processing on the historical behavior data to obtain standard user behavior data. A multi-dimensional behavioral feature construction module is used to extract behavioral features of the target user's feedback to each data transmission method based on the standard user behavior data, and to construct multi-dimensional behavioral features of the standard user behavior data based on the behavioral features. The reach probability calculation module is used to calculate the data transmission reach probability of each of the data transmission methods based on the multi-dimensional behavioral characteristics. The scene reach method calculation module is used to generate the scene reach method of the target user under different data transmission scenarios based on the data transmission reach probability. The switching strategy generation module is used to generate a data transmission method switching strategy corresponding to the target user using the scenario reach method.

[0007] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described data transmission mode switching method.

[0008] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described data transmission mode switching method.

[0009] In the aforementioned solution implemented by the data transmission mode switching method, system, computer equipment, and storage medium, data standardization processing of the target user's historical behavior data can yield more accurate and unified standard user behavior data, which is beneficial to improving the accuracy of subsequent user classification. By extracting behavioral characteristics of the target user's feedback to each data transmission mode based on the standard user behavior data, and constructing multi-dimensional behavioral features based on these characteristics, the target user's behavioral preferences for data transmission modes can be fully explored, further improving the accuracy of subsequent data transmission mode switching strategy generation. Calculating the data transmission reach probability of each data transmission mode based on the multi-dimensional behavioral features allows for more accurate calculation of the scene reach method in different data transmission scenarios. Generating the scene reach method for the target user in different data transmission scenarios based on the data transmission reach probability ensures the effectiveness of data transmission. Using the scene reach method to generate the data transmission mode switching strategy corresponding to the target user, the optimal data transmission mode can be determined based on the data transmission mode switching strategy, realizing dynamic decision-making for data push methods and effectively improving the flexibility of data transmission modes. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic diagram of an application environment for a data transmission mode switching method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a data transmission mode switching method in one embodiment of the present invention; Figure 3 This is a schematic diagram of a data transmission mode switching system according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention; Figure 5 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] The data transmission mode switching method provided in this embodiment of the invention can be applied to, for example... Figure 1 In this application environment, the client communicates with the server via a network. The server can obtain the target user's historical behavior data from the client, perform data standardization on the historical behavior data to obtain standard user behavior data; extract the target user's behavioral features for each data transmission method based on the standard user behavior data, construct multi-dimensional behavioral features of the standard user behavior data based on the behavioral features; calculate the data transmission reach probability corresponding to each data transmission method based on the multi-dimensional behavioral features; generate the target user's scenario reach method under different data transmission scenarios based on the data transmission reach probability; and generate a data transmission method switching strategy for the target user based on the scenario reach method. The optimal data transmission method for the target user can be determined based on the data transmission method switching strategy, realizing dynamic decision-making for data push methods and effectively improving the flexibility of data transmission methods. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a standalone server or a server cluster consisting of multiple servers. The invention will be described in detail below through specific embodiments.

[0014] Please see Figure 2 As shown, Figure 2 A flowchart illustrating a data transmission mode switching method provided in an embodiment of the present invention includes the following steps: S1. Obtain the target user's historical behavior data, and perform data standardization processing on the historical behavior data to obtain standard user behavior data.

[0015] In this embodiment of the invention, historical behavior data refers to user feedback data after push data is sent to users through different data transmission methods. For example, a telemarketing platform collects call records, including answer rate, rejection rate, call duration, and other data; an SMS platform obtains SMS sending records and read rates; user interaction data, such as message reading, reply frequency, and changes in friend relationships, are obtained from the WeChat Work (WeChat) system; and push notification reception rate and click rate are collected from the APP. User historical behavior data is feedback data collected when data is pushed to target users multiple times within a preset historical time period.

[0016] In this embodiment of the invention, the step of standardizing the historical behavior data to obtain standard user behavior data includes: The historical behavior data is classified to obtain numerical data and non-numerical data. The numerical data is deduplicated, outliers are removed, and missing values ​​are filled to obtain the target numerical data. The non-numerical data is converted to a new data format to obtain the target non-numerical data. By combining the target numerical data and the target non-numerical data, standard user behavior data corresponding to the historical behavior data is obtained.

[0017] In this embodiment of the invention, numerical data may include rating data, purchase frequency, dwell time, etc., while non-numerical data may include text feedback data, behavioral preference data, user interaction data, etc.

[0018] Furthermore, data deduplication can be achieved by removing duplicate values ​​through data distance, removing outliers through two-sided and one-sided test elimination, and detecting missing values ​​in the processed numerical data using the missmap function. If no missing values ​​are found, no further processing is performed. If missing values ​​are found, they can be filled using the mean of the data corresponding to the missing values ​​to obtain the target numerical data.

[0019] Preferably, data format conversion unifies the data format of non-numerical data, including: removing irrelevant characters from text, unifying encoding formats, unifying labels, unifying time formats, etc.

[0020] In this embodiment of the invention, by standardizing the historical behavior data of the target user, more accurate and uniform standard user behavior data can be obtained, which is beneficial to improving the accuracy of subsequent user classification.

[0021] S2. Extract the behavioral features of the target user's feedback on each data transmission method based on the standard user behavior data, and construct a multi-dimensional behavioral feature of the standard user behavior data based on the behavioral features.

[0022] In this embodiment of the invention, users can push information through various data transmission methods. Therefore, standard user behavior data contains feedback data of various data transmission methods. Behavioral features are the behavioral preferences of target users for data transmission methods. By analyzing behavioral features, we can capture the behavioral changes of target users for different data transmission methods, and thus better push information.

[0023] Specifically, the step of extracting the behavioral characteristics of the target user's feedback on each data transmission method based on the standard user behavior data includes: The system uses preset tagging rules to identify user feedback tags for each data transmission method in the standard user behavior data. Construct a feedback behavior feature vector for each of the data transmission methods based on the user feedback tags; Extract the context-aware features and self-attention features of the feedback behavior feature vector; The context-aware features and the self-attention features are multiplied element-wise to obtain the behavioral features corresponding to each data transmission method.

[0024] In this embodiment of the invention, the tagging rule is a business rule used to identify the user feedback tags corresponding to the data transmission methods in standard user behavior data. The user feedback tags are used to reflect the degree of positivity towards the data transmission method based on the feedback data. For example, the user feedback tags of the target user towards each data transmission method can be divided from high to low as highly positive, moderately positive, lowly negative, silent, and averse.

[0025] Furthermore, user feedback tags corresponding to standard user behavior data can be identified according to preset rules. For example, when the data transmission method is telemarketing, identification is based on the length of time the user answers the call in the feedback data; when the data transmission method is SMS, identification is based on the click-through rate of the SMS message; when the data transmission method is webpage, identification is based on the click-through time, etc. User feedback tag identification rules can be preset according to different data transmission methods, and further converted into executable tag rules for identifying user feedback tags in standard user behavior data. Through user feedback tags, the target user's preferences and responses to different data transmission methods can be determined. In detail, one-hot encoding can be used to convert user feedback tags into feature vectors. The feature vectors are then sorted by time to obtain feedback behavior feature vectors. Multilayer perceptron (MLP) is then used to extract contextual features from the feedback behavior feature vectors. This allows the behavioral logic relationships between user feedback tags to better understand the target user's tag conversion for data transmission methods.

[0026] Furthermore, the feature vectors converted from user feedback tags need to be transformed into dense low-dimensional embedding features. Each high-dimensional feature vector can be looked up in a pre-defined embedding table. The embedding table is the parameter matrix of the model, containing the embedding vectors of all feature vectors. The integer representation of each feature vector is used as the index of the embedding table to obtain the corresponding embedding vector, that is, to obtain the feedback behavior feature vector of the target user.

[0027] Specifically, a multilayer perceptron is a type of feedforward neural network consisting of hidden layers composed of multiple neurons. The context-aware features of the feedback behavior feature vector are obtained by weighted summation of the hidden layer neurons and calculation using activation functions.

[0028] Furthermore, self-attention calculation involves multiplying the feedback behavior feature vector with a learnable weight matrix to construct a query matrix (Q), a key matrix (K), and a value matrix (V). The dot product between the query vector and the key vector is calculated to obtain an attention score. An activation operation is performed on the attention score to obtain the probability distribution of the feature weights. The probability distribution is then multiplied with the value matrix to obtain the self-attention feature corresponding to the target user.

[0029] In this embodiment of the invention, by performing self-attention calculation, the relationship between different positions in the context-aware features can be considered in parallel, effectively capturing long-distance dependency information and extracting richer and more diverse features and relationships in the context-aware features, which is beneficial to improving the accuracy of subsequent multi-dimensional behavioral feature construction.

[0030] In this embodiment of the invention, multidimensional behavioral features are further extracted as better features among behavioral features, as well as high-order and low-order data of behavioral features.

[0031] Specifically, the step of constructing multidimensional behavioral features of the standard user behavior data based on the behavioral features includes: Multi-level cross-calculation is performed on the behavioral features to obtain cross features; Perform a second-order cross-calculation on the behavioral features to obtain low-order feature data; Multi-layer deep perception calculations are performed on the behavioral features to obtain high-order feature data; By combining the cross features, the low-order feature data, and the high-order feature data, multidimensional behavioral features of the standard user behavior data are obtained.

[0032] In this embodiment of the invention, parallel multi-layer cross network, FM (Factorization Machines) components, and DNN (Deep Neural Networks) models can be used to perform multi-layer cross computation, second-order cross computation, and multi-layer deep perception computation on behavioral features, respectively.

[0033] Among them, the multi-layer cross-computation network performs feature cross-computation on behavioral features through multiple cross-layers and then adds them to the input. It can effectively understand and capture explicit and bounded interactions between features. Furthermore, as the number of network layers increases, it can gradually increase the number of generated polynomial cross-features and simultaneously uncover higher-order nonlinear feature correlations.

[0034] The multi-level cross-calculation is described in detail below: in, Indicates behavioral characteristics, , … They represent the first layer, the second layer, and the third layer, respectively. The pre-defined weights of the layers, , … They represent the first layer, the second layer, and the third layer, respectively. The layer's preset bias parameters, This indicates the total number of layers in the multi-layer cross-computation. Indicates the first Cross features of layer output.

[0035] Furthermore, the FM (Factorization Machines) component can effectively model the second-order interactions between channel behavioral features. Specifically, it can first perform linear calculations on the behavioral features, then perform cross-calculations on the latent factor vectors of the behavioral features, and add the results of the two calculations together to capture the latent relationships and interaction patterns between the features in the behavioral features, thus obtaining low-order feature data. The second-order crossover can be calculated using the following formula: in, Represents low-order feature data, The first characteristic in behavioral traits One characteristic, Indicates the first Preset weight vectors corresponding to each feature This represents the total number of features in the behavioral characteristics. Indicates the first Features The corresponding latent factor vector, The first characteristic in behavioral traits Features The corresponding latent factor vector, express and The inner product between them.

[0036] In detail, multilayer perceptual computing is performed through a multilayer deep learning network, including an input layer, multiple hidden layers, and an output layer. The input layer receives behavioral features, the hidden layers are located between the input and output layers, each containing multiple neurons, and the output layer performs activation function operations to obtain high-order feature data.

[0037] In this embodiment of the invention, multidimensional behavioral features are obtained by extracting cross features, low-order feature data and high-order feature data in parallel. This can fully explore the behavioral preference features of target users for data transmission methods, thereby further improving the accuracy of subsequent data transmission method switching strategy generation.

[0038] S3. Calculate the data transmission reach probability corresponding to each of the data transmission methods based on the multidimensional behavioral characteristics.

[0039] In this embodiment of the invention, the data transmission reach probability is the probability of the user feedback tag after the target user selects the data transmission method. For example, the probability that the target user is highly positive, moderately positive, lowly negative, silent, or averse to data transmission method 1. The target user's preferences can be further analyzed through the data transmission reach probability.

[0040] Specifically, calculating the data transmission reach probability corresponding to each data transmission method based on the multi-dimensional behavioral features includes: The cross features, low-order feature data, and high-order feature data in the multidimensional behavioral features are standardized to obtain standard features; The standard features are concatenated to obtain the target user features; Perform a fully connected computation on the target user characteristics to obtain the data transmission reach probability corresponding to each of the data transmission methods.

[0041] In this embodiment of the invention, the standardization process makes the cross features, low-order feature data and high-order feature data have the same feature scale, so that they can be horizontally stacked to obtain the target user features.

[0042] Feature concatenation involves linearly concatenating cross features, low-order feature data, and high-order feature data from multi-dimensional behavioral features. Then, a preset activation function is used to map the target user features into the corresponding feature space. The probability value of user feedback tags for each data transmission method is calculated to obtain the data transmission reach probability.

[0043] In this embodiment of the invention, the data transmission reach probability is the user's preference for different data transmission methods. The data transmission reach probability can be used to calculate the target user's preference for different data transmission methods and response patterns.

[0044] For example, younger users pay more attention to social media platforms, while older users respond better to SMS and traditional TV and radio. With the popularization of smartphones and the development of technology, the behavior of older users in data transmission has begun to change from traditional TV and radio to social media platforms. At the same time, users with different interests and different spending power also respond differently to different data transmission methods. Therefore, the scenario reach probability can be used to more accurately calculate the scenario reach method in different data transmission scenarios.

[0045] S4. Generate the scenario reach method for the target user under different data transmission scenarios based on the data transmission reach probability.

[0046] In this embodiment of the invention, the scenario reach method under different data transmission scenarios is the optimal data transmission method for users under different data transmission scenarios. For example, the data transmission scenario when marketing data is pushed for the first time, the data transmission scenario when the effect of the first marketing data push is poor, and the data transmission scenario where the data transmission method needs to be switched due to failure or user inactivity.

[0047] In this embodiment of the invention, generating the scenario reach method for the target user under different data transmission scenarios based on the data transmission reach probability includes: The transmission method priority is calculated based on the data transmission reach probability. The data transmission methods are sorted according to their priority to obtain a priority sequence; Based on the priority sequence, data transmission scenarios are matched to obtain the scenario reach method for the target user under different data transmission scenarios.

[0048] In this embodiment of the invention, the target user's level of enthusiasm for the data transmission method is determined based on the data transmission reach probability. For example, the user feedback label corresponding to the highest data transmission method in data transmission reach probability 1 is a highly positive label, the user feedback label corresponding to the highest data transmission method in data transmission reach probability 2 is a moderately positive label, and the user feedback label corresponding to the highest data transmission method in data transmission reach probability 3 is a silent label. Then, the priority of the transmission methods in data transmission reach probabilities 1, 2, and 3 is sorted from high to low as data transmission reach probabilities 1, 2, and 3, thus obtaining a priority sequence.

[0049] In detail, priority sequences can be matched with preset data transmission scenarios according to pre-defined rules to obtain corresponding scenario-based delivery methods. For example, when pushing marketing data for the first time, the highest priority data transmission method in the priority sequence is selected for data push. When the user's level of engagement in real-time feedback data is lower than that of low engagement, and a switch to the data transmission method is required, the next priority data transmission method is selected for data push. This process is repeated to obtain scenario-based delivery methods under different data transmission scenarios, thus ensuring the effectiveness of data push.

[0050] S5. Generate a data transmission method switching strategy for the target user using the scenario reach method.

[0051] In this embodiment of the invention, the data transmission method switching strategy is a strategy that flexibly switches and combines the data transmission methods of the target user. For example, after a phone push to the target user fails or succeeds, the data transmission method corresponding to the telephone marketing is switched to ensure that the target user can receive the push data in the most efficient way.

[0052] In this embodiment of the invention, generating the data transmission method switching strategy corresponding to the target user using the scenario reach method includes: A transmission method rule engine is constructed based on the described scenario reach method; Collect real-time user feedback data from the target users; The real-time data transmission scenario corresponding to the target user is determined based on the real-time user feedback data. Based on the real-time data transmission scenario, the data transmission mode switching strategy is generated using the transmission mode rule engine.

[0053] In this embodiment of the invention, the transmission mode rule engine converts the transmission mode priority into executable business rules. Based on the transmission mode rule engine, the optimal data transmission mode is determined, and dynamic decision-making is achieved, which can improve the flexibility of data transmission mode.

[0054] Specifically, real-time user feedback data refers to the real-time feedback data from target users on the data-driven data transmission method. For example, when the real-time data transmission method is telemarketing, the real-time user feedback data may be user activity level and call duration. When the real-time data transmission method is SMS push, the real-time user feedback data may be unsubscription rate and click-through rate, etc. Different data transmission methods collect corresponding user feedback data to obtain real-time user feedback data.

[0055] Furthermore, the data transmission method is determined according to the priority of the transmission method in different data transmission scenarios. That is, the data transmission method with the highest priority is used first to push data to the user. If it fails or the user is inactive, the data transmission method corresponding to the next priority is switched. If it succeeds or the user is active, the data transmission method of the previous one is used to push the data. In this way, a transmission method rule engine is built, and the data transmission method rule engine determines the data transmission method selected by the target user in different data transmission scenarios.

[0056] Furthermore, real-time data transmission scenarios can be determined based on real-time user feedback data, i.e., user feedback tags for data transmission methods. Data transmission methods are matched according to real-time data transmission scenarios. For example, when the user activity level in real-time user feedback data is lower than low level, it is necessary to switch the data transmission method. The data transmission method switching strategy corresponding to the target user is obtained according to the business rules in the transmission method rule engine, so as to flexibly push data.

[0057] In this embodiment of the invention, the data transmission mode switching strategy can determine the optimal data transmission mode, realize dynamic decision-making on data push mode, and effectively improve the flexibility of data transmission mode.

[0058] As can be seen, in the above scheme, by standardizing the historical behavior data of target users, more accurate and uniform standard user behavior data can be obtained, which is conducive to improving the accuracy of subsequent user classification. Extracting behavioral characteristics of target users' responses to each data transmission method based on the standard user behavior data, and constructing multi-dimensional behavioral features based on these characteristics, can fully explore the target users' behavioral preferences for data transmission methods, further improving the accuracy of subsequent data transmission method switching strategy generation. Calculating the data transmission reach probability of each data transmission method based on the multi-dimensional behavioral features allows for more accurate calculation of the scene reach method in different data transmission scenarios. Generating the scene reach method for target users in different data transmission scenarios based on the data transmission reach probability ensures the effectiveness of data transmission. Using the scene reach method to generate the data transmission method switching strategy corresponding to the target user, the optimal data transmission method can be determined based on the data transmission method switching strategy, realizing dynamic decision-making on data push methods and effectively improving the flexibility of data transmission methods.

[0059] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0060] In one embodiment, a data transmission mode switching system is provided, which corresponds one-to-one with the data transmission mode switching methods described in the above embodiments. For example... Figure 3 As shown, the data transmission mode switching system includes a data processing module 101, a multi-dimensional behavioral feature construction module 102, a reach probability calculation module 103, a scene reach method calculation module 104, and a switching strategy generation module 105. Detailed descriptions of each functional module are as follows: Data processing module 101 is used to acquire historical behavior data of target users, and perform data standardization processing on the historical behavior data to obtain standard user behavior data; The multidimensional behavioral feature construction module 102 is used to extract the behavioral features of the target user's feedback to each data transmission method based on the standard user behavior data, and to construct the multidimensional behavioral features of the standard user behavior data based on the behavioral features. The reach probability calculation module 103 is used to calculate the data transmission reach probability of each of the data transmission methods based on the multi-dimensional behavioral characteristics; The scene reach method calculation module 104 is used to generate the scene reach method of the target user under different data transmission scenarios based on the data transmission reach probability. The switching strategy generation module 105 is used to generate a data transmission mode switching strategy corresponding to the target user using the scenario reach method.

[0061] In one embodiment, when the data processing module 101 performs data standardization processing on the historical behavior data to obtain standard user behavior data, it is used to: The historical behavior data is classified to obtain numerical data and non-numerical data. The numerical data is deduplicated, outliers are removed, and missing values ​​are filled to obtain the target numerical data. The non-numerical data is converted to a new data format to obtain the target non-numerical data. By combining the target numerical data and the target non-numerical data, standard user behavior data corresponding to the historical behavior data is obtained.

[0062] In one embodiment, the multi-dimensional behavioral feature construction module 102, when extracting the behavioral features of the target user's feedback on each data transmission method based on the standard user behavior data, is used to: The system uses preset tagging rules to identify user feedback tags for each data transmission method in the standard user behavior data. Construct a feedback behavior feature vector for each of the data transmission methods based on the user feedback tags; Extract the context-aware features and self-attention features of the feedback behavior feature vector; The context-aware features and the self-attention features are multiplied element-wise to obtain the behavioral features corresponding to each data transmission method.

[0063] In one embodiment, when constructing multidimensional behavioral features of the standard user behavior data based on the channel behavioral features, the multidimensional behavioral feature construction module 102 is used to: Multi-level cross-calculation is performed on the behavioral features to obtain cross features; Perform a second-order cross-calculation on the behavioral features to obtain low-order feature data; Multi-layer deep perception calculations are performed on the behavioral features to obtain high-order feature data; By combining the cross features, the low-order feature data, and the high-order feature data, multidimensional behavioral features of the standard user behavior data are obtained.

[0064] In one embodiment, when calculating the data transmission reach probability corresponding to each of the data transmission methods based on the multi-dimensional behavioral characteristics, the reach probability calculation module 103 is used to: The cross features, low-order feature data, and high-order feature data in the multidimensional behavioral features are standardized to obtain standard features; The standard features are concatenated to obtain the target user features; Perform a fully connected computation on the target user characteristics to obtain the data transmission reach probability corresponding to each of the data transmission methods.

[0065] In one embodiment, when the scene reach method calculation module 104 generates the scene reach method for the target user under different data transmission scenarios based on the data transmission reach probability, it is used to: The transmission method priority is calculated based on the data transmission reach probability. The data transmission methods are sorted according to their priority to obtain a priority sequence; Based on the priority sequence, data transmission scenarios are matched to obtain the scenario reach method for the target user under different data transmission scenarios.

[0066] In one embodiment, when the switching policy generation module 105 generates a data transmission method switching policy corresponding to the target user using the data transmission method, it is used to: A transmission method rule engine is constructed based on the described scenario reach method; Collect real-time user feedback data from the target users; The real-time data transmission scenario corresponding to the target user is determined based on the real-time user feedback data. Based on the real-time data transmission scenario, the data transmission mode switching strategy is generated using the transmission mode rule engine.

[0067] This invention provides a data transmission mode switching system. By standardizing the historical behavior data of target users, more accurate and uniform standard user behavior data can be obtained, which is beneficial to improving the accuracy of subsequent user classification. Based on the standard user behavior data, behavioral features of target users' responses to each data transmission mode are extracted. Multi-dimensional behavioral features are constructed based on these features to fully explore the target users' behavioral preferences for data transmission modes, further improving the accuracy of subsequent data transmission mode switching strategy generation. The data transmission reach probability of each data transmission mode is calculated based on the multi-dimensional behavioral features, allowing for more accurate calculation of the scene reach method in different data transmission scenarios. Generating scene reach methods for target users in different data transmission scenarios based on the data transmission reach probability ensures the effectiveness of data transmission. Using the scene reach methods to generate the corresponding data transmission mode switching strategy for the target user, the optimal data transmission mode can be determined based on the data transmission mode switching strategy, realizing dynamic decision-making for data push methods and effectively improving the flexibility of data transmission modes.

[0068] Specific limitations regarding the data transmission mode switching system can be found in the limitations of the data transmission mode switching method described above, and will not be repeated here. Each module in the aforementioned data transmission mode switching system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in the computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0069] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a data transmission mode switching method on the server side.

[0070] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, memory, network interface, display screen, and input system connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements a data transmission mode switching method on the client side, fulfilling its functions or steps. In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: Obtain historical behavior data of the target user, and perform data standardization processing on the historical behavior data to obtain standard user behavior data; Based on the standard user behavior data, extract the behavioral features of the target user's feedback to each data transmission method, and construct a multidimensional behavioral feature of the standard user behavior data based on the behavioral features. Calculate the data transmission reach probability corresponding to each of the data transmission methods based on the multidimensional behavioral characteristics; Based on the data transmission reach probability, the scenario reach method for the target user under different data transmission scenarios is generated; The data transmission method switching strategy corresponding to the target user is generated using the scenario-based outreach method.

[0071] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Obtain historical behavior data of the target user, and perform data standardization processing on the historical behavior data to obtain standard user behavior data; Based on the standard user behavior data, extract the behavioral features of the target user's feedback to each data transmission method, and construct a multidimensional behavioral feature of the standard user behavior data based on the behavioral features. Calculate the data transmission reach probability corresponding to each of the data transmission methods based on the multidimensional behavioral characteristics; Based on the data transmission reach probability, the scenario reach method for the target user under different data transmission scenarios is generated; The data transmission method switching strategy corresponding to the target user is generated using the scenario-based outreach method.

[0072] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0073] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0074] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

[0075] Finally, it should be noted that if any software tools or components not belonging to this company appear in the embodiments of the application, they are merely illustrative examples and do not represent actual use. The embodiments described above are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for switching data transmission modes, characterized in that, include: Obtain historical behavior data of the target user, and perform data standardization processing on the historical behavior data to obtain standard user behavior data; Based on the standard user behavior data, extract the behavioral features of the target user's feedback to each data transmission method, and construct a multidimensional behavioral feature of the standard user behavior data based on the behavioral features. Calculate the data transmission reach probability corresponding to each of the data transmission methods based on the multidimensional behavioral characteristics; Based on the data transmission reach probability, the scenario reach method for the target user under different data transmission scenarios is generated; The data transmission method switching strategy corresponding to the target user is generated using the scenario-based outreach method.

2. The data transmission mode switching method as described in claim 1, characterized in that, The process of standardizing the historical behavior data to obtain standard user behavior data includes: The historical behavior data is classified to obtain numerical data and non-numerical data. The numerical data is deduplicated, outliers are removed, and missing values ​​are filled to obtain the target numerical data. The non-numerical data is converted to a new data format to obtain the target non-numerical data. By combining the target numerical data and the target non-numerical data, standard user behavior data corresponding to the historical behavior data is obtained.

3. The data transmission mode switching method as described in claim 1, characterized in that, The step of extracting the behavioral features of the target user's feedback on each data transmission method based on the standard user behavior data includes: The system uses preset tagging rules to identify user feedback tags for each data transmission method in the standard user behavior data. Construct a feedback behavior feature vector for each of the data transmission methods based on the user feedback tags; Extract the context-aware features and self-attention features of the feedback behavior feature vector; The context-aware features and the self-attention features are multiplied element-wise to obtain the behavioral features corresponding to each data transmission method.

4. The data transmission mode switching method as described in claim 1, characterized in that, The construction of multidimensional behavioral features of the standard user behavior data based on the behavioral features includes: Multi-level cross-calculation is performed on the behavioral features to obtain cross features; Perform a second-order cross-calculation on the behavioral features to obtain low-order feature data; Multi-layer deep perception calculations are performed on the behavioral features to obtain high-order feature data; By combining the cross features, the low-order feature data, and the high-order feature data, multidimensional behavioral features of the standard user behavior data are obtained.

5. The data transmission mode switching method as described in claim 1, characterized in that, The step of calculating the data transmission reach probability corresponding to each data transmission method based on the multi-dimensional behavioral features includes: The cross features, low-order feature data, and high-order feature data in the multidimensional behavioral features are standardized to obtain standard features; The standard features are concatenated to obtain the target user features; Perform a fully connected computation on the target user characteristics to obtain the data transmission reach probability corresponding to each of the data transmission methods.

6. The data transmission mode switching method as described in claim 1, characterized in that, The step of generating the scenario-based outreach method for the target user under different data transmission scenarios based on the data transmission outreach probability includes: The transmission method priority is calculated based on the data transmission reach probability. The data transmission methods are sorted according to their priority to obtain a priority sequence; Based on the priority sequence, data transmission scenarios are matched to obtain the scenario reach method for the target user under different data transmission scenarios.

7. The data transmission mode switching method as described in claim 1, characterized in that, The method of generating a data transmission mode switching strategy corresponding to the target user using the scenario-based outreach method includes: A transmission method rule engine is constructed based on the described scenario reach method; Collect real-time user feedback data from the target users; The real-time data transmission scenario corresponding to the target user is determined based on the real-time user feedback data. Based on the real-time data transmission scenario, the data transmission mode switching strategy is generated using the transmission mode rule engine.

8. A data transmission mode switching system, characterized in that, include: The data processing module is used to acquire the historical behavior data of the target user, and to perform data standardization processing on the historical behavior data to obtain standard user behavior data. A multi-dimensional behavioral feature construction module is used to extract behavioral features of the target user's feedback to each data transmission method based on the standard user behavior data, and to construct multi-dimensional behavioral features of the standard user behavior data based on the behavioral features. The reach probability calculation module is used to calculate the data transmission reach probability corresponding to each of the data transmission methods based on the multi-dimensional behavioral characteristics. The scene reach method calculation module is used to generate the scene reach method of the target user under different data transmission scenarios based on the data transmission reach probability. The switching strategy generation module is used to generate a data transmission method switching strategy corresponding to the target user using the scenario reach method.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the data transmission mode switching method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the data transmission mode switching method as described in any one of claims 1 to 7.

Images