Financial knowledge information generation method and device, and program product

Abstract

The invention discloses a financial knowledge information generation method and device and a program product, relates to the field of artificial intelligence, and is applied to the financial science and technology field, and the method comprises the steps: collecting user behavior data of a target user on a financial education platform under the condition of obtaining the authorization of the target user, and carrying out the feature extraction of the user behavior data, obtaining a first behavior feature; inputting the first behavior feature into a target evaluation model, and outputting a target score; extracting a second behavior feature from the user behavior data, the second behavior feature representing a feature of a response behavior of the target user for the answer error; and generating financial knowledge information to be learned in the financial education platform according to the first behavior feature, the second behavior feature and the target score. Through application of the method and the device, the problem that the user participation degree in the financial education platform is relatively low due to the fact that learning contents in the financial education platform are generally designed in a unified manner and the consideration for different user requirements and different learning behaviors is lacked in the prior art is solved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence and is applied to the field of financial technology. Specifically, it relates to a method, device, and program product for generating financial knowledge information. Background Technology

[0002] In the current field of financial education, traditional educational platforms mostly adopt standardized teaching content and models, failing to fully consider the differences in individual users' knowledge base, learning habits, and psychological inclinations. This one-size-fits-all educational strategy often results in low user engagement and inefficient knowledge absorption, especially when faced with complex financial concepts and decision-making, lacking targeted guidance and incentives. Furthermore, with the deepening of digital transformation, the collection and analysis of user data has become an important means to improve educational effectiveness, but it has also raised concerns about privacy protection. Current technologies, with their centralized data processing methods, not only increase the risk of information leakage but may also violate relevant regulations by neglecting user privacy, affecting user trust.

[0003] Currently, there is no effective solution to the problem that the learning content in financial education platforms is usually uniformly designed, lacking consideration for different user needs and learning behaviors, resulting in low user participation. Summary of the Invention

[0004] The main purpose of this application is to provide a method, device, and program product for generating financial knowledge information, in order to solve the problem that the learning content in financial education platforms is usually uniformly designed and lacks consideration for different user needs and different learning behaviors, resulting in low user participation in financial education platforms.

[0005] To achieve the above objectives, according to one aspect of this application, a method for generating financial knowledge information is provided. The method includes: collecting user behavior data of the target user on a financial education platform with the user's authorization, and extracting features from the user behavior data to obtain first behavioral features, wherein the first behavioral features represent the target user's learning behavior characteristics on the financial education platform; inputting the first behavioral features into a target evaluation model and outputting a target score, wherein the target score includes at least a first score and a second score, the first score representing the target user's preference information for reward timeliness, and the second score representing the target user's mastery of financial knowledge, the target evaluation model being obtained by training a preset model based on a federated learning algorithm using historical user learning behavior on the financial education platform; extracting second behavioral features from the user behavior data, wherein the second behavioral features represent the characteristics of the target user's response behavior to incorrect answers; and generating financial knowledge information to be learned from the financial education platform based on the first behavioral features, the second behavioral features, and the target score.

[0006] Furthermore, before inputting the first behavioral feature into the target evaluation model and outputting the target score, the above method also includes: preprocessing historical learning behavior data of historical users to obtain historical behavioral features, and constructing a training set based on the historical behavioral features and historical target scores; encrypting the training set, training a local preset model using the encrypted training set, encrypting the trained model parameters to obtain encrypted parameters; uploading the encrypted parameters to the server and receiving fusion parameters from the server, wherein the server decrypts the encrypted parameters uploaded by each client and fuses the decrypted model parameters to obtain fusion parameters; updating the model parameters of the preset model based on the fusion parameters, training the updated preset model using the encrypted training set, uploading the encrypted model parameters back to the server, iteratively training the preset model based on the fusion parameters from the server until the model converges to obtain the target evaluation model.

[0007] Further, extracting a second behavioral feature from user behavior data includes: extracting a first feature vector and a second feature vector from user behavior data using a feature extraction model, wherein the first feature vector represents the target user's answer errors and the second feature vector represents the target user's response behavior information in response to answer errors; adjusting the initial behavioral coefficients based on the first behavioral feature to obtain target behavioral coefficients; and generating the second behavioral feature based on the first feature vector, the second feature vector, and the target behavioral coefficient.

[0008] Furthermore, based on the first behavioral characteristic, the second behavioral characteristic, and the target score, financial knowledge information to be learned in the financial education platform is generated, including: determining learning effectiveness indicators based on user behavior data, wherein the learning effectiveness indicators include at least: a first indicator, a second indicator, and a third indicator, where the first indicator represents the target user's correct answer rate, the second indicator represents the target user's understanding of financial knowledge, and the third indicator represents the target user's course completion level in the financial education platform; determining a candidate knowledge set based on the financial knowledge information in the financial education platform; calculating the score distribution of the candidate knowledge information in the candidate knowledge set based on the candidate knowledge set, the learning effectiveness indicators, the first behavioral characteristic, the second behavioral characteristic, and the target score; and filtering the candidate knowledge information in the candidate knowledge set based on the score distribution to obtain the financial knowledge information to be learned.

[0009] Further, the score distribution of candidate knowledge information in the candidate knowledge set is calculated based on the candidate knowledge set, learning effect indicators, first behavioral features, second behavioral features, and target scores. This includes: determining a first score based on candidate knowledge information, learning effect indicators, and a first function, wherein the first function is used to calculate the degree of matching between the learning difficulty of the candidate knowledge information and the knowledge already mastered by the target user; determining a second score based on candidate knowledge information, learning behavioral features, and a second function, wherein the learning behavioral features include at least one of the following: target score, second behavioral features, and the second function is used to calculate the degree of matching between the candidate knowledge information and the learning behavioral features; determining a third score based on candidate knowledge information, first behavioral features, and a third function, wherein the third function is used to calculate the degree of interest of the target user in the candidate knowledge information; and calculating the score distribution of candidate knowledge information based on the first score, second score, and third score.

[0010] Furthermore, after generating the financial knowledge information to be learned in the financial education platform based on the first behavioral characteristics, the second behavioral characteristics, and the target score, the above method also includes: determining the number of target knowledge points already learned by the target user based on user behavior data, and determining the course completion progress based on the number of target knowledge points; determining the score information for each knowledge point based on user behavior data, and calculating the average score based on the score information for each knowledge point; determining the learning time of the target user based on user behavior data, and determining the level of enthusiasm of the target user's learning behavior; determining the display method of the financial knowledge information to be learned based on the course completion progress, average score, learning time, enthusiasm, second behavioral characteristics, and target score, and pushing the financial knowledge information to be learned to the target user according to the display method.

[0011] Furthermore, after pushing the financial knowledge information to be learned to the target user according to the display method, the above method also includes: after detecting that the target user has completed the learning process of the financial knowledge information to be learned, collecting the target user's transaction information; inputting the target user's transaction information and the financial knowledge information already learned by the target user into the target recommendation model, and outputting the financial knowledge recommendation information, wherein the target recommendation model is a model trained on a preset model using historical users' transaction information and historical users' financial knowledge information based on a federated learning algorithm.

[0012] Furthermore, user behavior data of target users on the financial education platform is collected, and feature extraction is performed on the user behavior data to obtain the first behavioral feature. This includes: obtaining user information of the target users, wherein the user information includes at least one of the following: age information, educational background information, occupation, and financial knowledge level, wherein the financial knowledge level is determined based on a financial knowledge test questionnaire; collecting user behavior data generated by the target users on various clients of the financial education platform, wherein the user behavior data includes at least one of the following: learning duration, learning completion rate, click data, learning feedback data, and transaction behavior information; performing data preprocessing operations on the user behavior data, and performing feature extraction operations on the processed user behavior data to obtain the first behavioral feature.

[0013] To achieve the above objectives, according to another aspect of this application, a financial knowledge information generation apparatus is provided. The apparatus includes: a first acquisition unit, configured to, with the authorization of a target user, acquire user behavior data of the target user on a financial education platform, and extract features from the user behavior data to obtain first behavioral features, wherein the first behavioral features represent the learning behavior characteristics of the target user on the financial education platform; an evaluation unit, configured to input the first behavioral features into a target evaluation model and output a target score, wherein the target score includes at least: a first score and a second score, the first score representing the target user's preference information for reward timeliness, and the second score representing the target user's mastery of financial knowledge, the target evaluation model being obtained by training a preset model based on a federated learning algorithm according to the historical user learning behavior on the financial education platform; an extraction unit, configured to extract second behavioral features from the user behavior data, wherein the second behavioral features represent the characteristics of the target user's response behavior to incorrect answers; and a generation unit, configured to generate financial knowledge information to be learned from the financial education platform based on the first behavioral features, the second behavioral features, and the target score.

[0014] Furthermore, the device also includes: a construction unit, used to preprocess historical learning behavior data of historical users to obtain historical behavior features before inputting the first behavioral features into the target evaluation model and outputting the target score, and to construct a training set based on the historical behavioral features and historical target scores; a first training unit, used to encrypt the training set, train a local preset model using the encrypted training set, and encrypt the trained model parameters to obtain encrypted parameters; a receiving unit, used to upload the encrypted parameters to the server and receive the fusion parameters sent by the server, wherein the server decrypts the encrypted parameters uploaded by each client and fuses the decrypted model parameters to obtain fusion parameters; a second training unit, used to update the model parameters of the preset model according to the fusion parameters, train the updated preset model using the encrypted training set, upload the encrypted model parameters back to the server, and iteratively train the preset model based on the fusion parameters sent by the server until the model converges to obtain the target evaluation model.

[0015] Furthermore, the extraction unit includes: an extraction subunit, used to extract a first feature vector and a second feature vector from user behavior data through a feature extraction model, wherein the first feature vector represents the target user's answer error situation, and the second feature vector represents the target user's response behavior information in response to answer errors; an adjustment subunit, used to adjust the initial behavior coefficients according to the first behavior feature to obtain target behavior coefficients; and a generation subunit, used to generate the second behavior feature based on the first feature vector, the second feature vector, and the target behavior coefficient.

[0016] Furthermore, the generation unit includes: a first determining subunit, used to determine learning effectiveness indicators based on user behavior data, wherein the learning effectiveness indicators include at least: a first indicator, a second indicator, and a third indicator, wherein the first indicator represents the target user's correct answer rate, the second indicator represents the target user's understanding of financial knowledge, and the third indicator represents the target user's course completion level on the financial education platform; a second determining subunit, used to determine a candidate knowledge set based on financial knowledge information from the financial education platform; a calculation subunit, used to calculate the score distribution of candidate knowledge information in the candidate knowledge set based on the candidate knowledge set, learning effectiveness indicators, the first behavioral feature, the second behavioral feature, and the target score; and a filtering subunit, used to filter candidate knowledge information from the candidate knowledge set based on the score distribution of candidate knowledge information to obtain the financial knowledge information to be learned.

[0017] Further, the calculation subunit includes: a first determining module, used to determine a first score based on candidate knowledge information, learning effect indicators, and a first function, wherein the first function is used to calculate the degree of matching between the learning difficulty of the candidate knowledge information and the knowledge already mastered by the target user; a second determining module, used to determine a second score based on candidate knowledge information, learning behavior characteristics, and a second function, wherein the learning behavior characteristics include at least one of the following: target score, second behavior characteristics, and the second function is used to calculate the degree of matching between the candidate knowledge information and the learning behavior characteristics; a third determining module, used to determine a third score based on candidate knowledge information, the first behavior characteristics, and a third function, wherein the third function is used to calculate the degree of interest of the target user in the candidate knowledge information; and a calculation module, used to calculate the score distribution of the candidate knowledge information based on the first score, the second score, and the third score.

[0018] Furthermore, the device also includes: a first determining unit, used to determine the number of target knowledge points already learned by the target user based on user behavior data after generating financial knowledge information to be learned in the financial education platform based on the first behavioral characteristics, the second behavioral characteristics, and the target score, and to determine the course completion progress based on the number of target knowledge points; a calculation unit, used to determine the score information of each knowledge point based on user behavior data, and to calculate the average score based on the score information of each knowledge point; a second determining unit, used to determine the learning time of the target user based on user behavior data, and to determine the level of enthusiasm of the target user's learning behavior; and a push unit, used to determine the display method of the financial knowledge information to be learned based on the course completion progress, average score, learning time, enthusiasm, second behavioral characteristics, and target score, and to push the financial knowledge information to be learned to the target user according to the display method.

[0019] Furthermore, the device also includes: a second acquisition unit, used to acquire the target user's transaction information after the target user has completed the learning process of the financial knowledge information to be learned after the target user pushes the financial knowledge information to be learned to the target user according to the display method; and a processing unit, used to input the target user's transaction information and the financial knowledge information already learned by the target user into the target recommendation model and output the financial knowledge recommendation information, wherein the target recommendation model is a model trained on a preset model using historical users' transaction information and historical users' financial knowledge information based on a federated learning algorithm.

[0020] Furthermore, the first collection unit includes: an acquisition unit for acquiring user information of the target user, wherein the user information includes at least one of the following: age information, educational background information, occupation, and financial knowledge level, wherein the financial knowledge level is determined based on a financial knowledge test questionnaire; a third collection unit for collecting user behavior data generated by the target user on various clients of the financial education platform, wherein the user behavior data includes at least one of the following: learning duration, learning completion rate, click data, learning feedback data, and transaction behavior information; and a processing unit for performing data preprocessing operations on the user behavior data and performing feature extraction operations on the processed user behavior data to obtain the first behavioral feature.

[0021] To achieve the above objectives, according to one aspect of this application, a computer program product is provided, including a computer program that, when executed by a processor, implements any of the above-mentioned methods for generating financial knowledge information, and when executed by a processor, implements the steps of the methods for generating financial knowledge information in various embodiments of this application.

[0022] To achieve the above objectives, according to one aspect of this application, a computer-readable storage medium is provided, comprising stored computer instructions, wherein, when the computer instructions are executed by a processor, the method for generating any of the aforementioned financial knowledge information is implemented.

[0023] To achieve the above objectives, according to one aspect of this application, an electronic device is provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement any of the above-described methods for generating financial knowledge information.

[0024] In this embodiment, user behavior data of the target user on a financial education platform is collected with the user's authorization, and features are extracted from the user behavior data to obtain a first behavioral feature, which represents the target user's learning behavior characteristics on the financial education platform. The first behavioral feature is input into a target evaluation model to output a target score, which includes at least a first score and a second score. The first score represents the target user's preference for reward timeliness, and the second score represents the target user's mastery of financial knowledge. The target evaluation model is obtained by training a preset model based on a federated learning algorithm according to the historical user learning behavior on the financial education platform. A second behavioral feature is extracted from the user behavior data, which represents the target user's response behavior to incorrect answers. Based on the first behavioral feature, the second behavioral feature, and the target score, financial knowledge information to be learned on the financial education platform is generated. This solves the technical problem that the learning content on financial education platforms is usually uniformly designed, lacking consideration for different user needs and different learning behaviors, resulting in low user participation on financial education platforms.

[0025] By inputting the first behavioral feature into a target evaluation model trained by a federated learning algorithm, a target score encompassing the user's reward timeliness preference and financial knowledge mastery can be output, achieving personalized educational needs assessment and further optimizing the allocation of educational resources. Furthermore, by extracting features from the user's response behavior to incorrect answers, a second behavioral feature is formed, which helps identify the user's learning difficulties and areas for improvement. The financial education platform's learning content, generated based on the first and second behavioral features and the target score, can highly match the user's learning ability and preferences, achieving personalized recommendations of educational content, improving user learning efficiency and satisfaction, and ultimately increasing user engagement on the financial education platform. Attached Figure Description

[0026] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0027] Figure 1 This is a hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for generating financial knowledge information, according to Embodiment 1 of this application;

[0028] Figure 2 This is a flowchart of an optional method for generating financial knowledge information according to Embodiment 1 of this application;

[0029] Figure 3This is a schematic diagram of the optional process for generating financial knowledge information to be recommended in a financial education platform, according to Embodiment 1 of this application.

[0030] Figure 4 This is a schematic diagram of a financial knowledge information generation device according to Embodiment 2 of this application;

[0031] Figure 5 This is a schematic diagram of an electronic device for generating financial knowledge information according to Embodiment 3 of this application. Detailed Implementation

[0032] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0033] It should be noted that the processing methods, apparatus, storage media, and electronic devices specified in this application can be used in the fintech field to improve user participation during the use of financial education platforms, and can also be used in any field other than fintech. The application fields of the processing methods, apparatus, storage media, and electronic devices specified in this application are not limited.

[0034] It should be noted that the information collected in this application (including but not limited to user device information, user personal information, collected data, used data, generated data, processed data, etc.) and the data (including but not limited to data used for analysis, stored data, displayed data, collected information, used information, generated information, processed information, etc.) are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, have taken necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse. For example, this system has interfaces with relevant users or organizations, providing users with corresponding operation entry points for users to choose to agree to or refuse automated decision results; if the user chooses to refuse, the process proceeds to the expert decision-making stage.

[0035] Example 1

[0036] According to an embodiment of this application, a method embodiment for generating financial knowledge information is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0037] The method embodiment provided in Embodiment 1 of this application can be executed in a mobile terminal, computer terminal or similar computing device. Figure 1 This is a hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for generating financial knowledge information, according to Embodiment 1 of this application. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0038] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0039] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the financial knowledge information generation method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned financial knowledge information generation method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0040] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0041] The display may be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0042] Under the aforementioned operating environment, this application provides the following: Figure 2 The method for generating financial knowledge information is shown. Figure 2 This is a flowchart of an optional method for generating financial knowledge information according to Embodiment 1 of this application.

[0043] Step S201: With the authorization of the target user, collect the user behavior data of the target user on the financial education platform, and extract features from the user behavior data to obtain the first behavior feature, wherein the first behavior feature represents the learning behavior characteristics of the target user on the financial education platform.

[0044] In this embodiment 1, to automatically generate personalized financial knowledge for target users on the financial education platform, it is necessary to collect their interaction records on the platform, including but not limited to learning duration, course completion status, and test results. Subsequently, data analysis techniques are used to extract key behavioral features from this raw behavioral data, forming a first behavioral feature set. Through this step, the platform can obtain a representation of the user's individualized learning behavior, providing a data foundation for further providing customized educational content and services.

[0045] It is important to note that in this application, before collecting any user data, a detailed data usage policy will be presented to the user, including the data types, collection purposes, processing methods, and protective measures, in a transparent manner. The acquisition of user behavioral information strictly adheres to user privacy protection principles, ensuring that it is done only after the target user has given explicit authorization. Furthermore, when designing and implementing data collection strategies, the principle of data minimization will be fully considered, collecting only user behavioral data directly related to educational services, avoiding the collection of unnecessary personal information, and thus reducing privacy risks.

[0046] Step S202: Input the first behavioral feature into the target evaluation model and output the target score. The target score includes at least a first score and a second score. The first score represents the target user's preference information for reward timeliness, and the second score represents the target user's mastery of financial knowledge. The target evaluation model is obtained by training a preset model based on the historical user learning behavior of the financial education platform using a federated learning algorithm.

[0047] In this embodiment 1, the extracted first behavioral feature is used as input to a target evaluation model trained by a federated learning algorithm. This model is built based on historical user data from the financial education platform. After computation, the model outputs a target score, which is divided into two parts: the first score reflects the target user's preference for reward timeliness, quantifying the user's tendency towards immediate or delayed rewards; the second score assesses the user's actual mastery of financial knowledge, providing a quantitative indicator of the user's current knowledge level. This evaluation process relies on the federated learning mechanism to ensure the security of user privacy during data processing. In this way, the platform can accurately obtain users' learning characteristics and knowledge mastery status, providing crucial information for the design of subsequent personalized education programs.

[0048] For example, the input to the target evaluation model is user behavior data held by each client (user's local device or bank branch's local device), which can be represented as: ,in, Let represent the data vector of the i-th user, containing m features, including but not limited to: static features, such as user age, education background, occupation, and financial literacy level (survey questionnaire or initial test results); and dynamic features, such as learning behavior (learning duration, completion rate, click habits), consumption behavior (frequency of financial product use, transaction amount), and feedback data (satisfaction rating). The output of the target evaluation model includes: β, which is the first score mentioned above, and γ, which is the second score mentioned above.

[0049] The first score represents a user's preference for future feedback, a method used to measure a user's preference between immediate and delayed rewards. For example, individuals with lower beta tend to be short-sighted, valuing immediate gains or satisfaction more and showing less motivation to learn about potentially greater future gains. Conversely, individuals with higher beta have a greater motivation to learn about future rewards. Overconfidence, another important concept in behavioral economics, refers to an individual's excessive confidence in the accuracy of their judgment, thus underestimating uncertainty in decision-making. The second score refers to a user's confidence in their own financial knowledge.

[0050] Step S203: Extract the second behavioral feature from the user behavior data, wherein the second behavioral feature represents the characteristics of the target user's response behavior to incorrect answers.

[0051] In this embodiment 1, a second behavioral feature is extracted from the collected user behavior data through data processing and analysis. This feature focuses on describing the degree to which negative emotions influence user learning behavior during the learning process. By extracting features from these specific behavioral data, it is possible to reveal users' self-regulation behaviors when facing knowledge gaps or comprehension obstacles, thereby providing a detailed basis for the formulation of educational strategies.

[0052] Step S204: Generate financial knowledge information to be learned in the financial education platform based on the first behavioral feature, the second behavioral feature, and the target score.

[0053] In this embodiment 1, based on the first behavioral characteristic—the learning behavior attributes of the target user—and the second behavioral characteristic—the user's specific reaction to incorrect answers—and combined with the target user's preference for reward timeliness and level of financial knowledge quantified in the target score, financial knowledge information to be learned on the financial education platform is generated. This generation process considers the user's learning habits, knowledge base, and how they handle errors, aiming to accurately match user needs and tailor learning content to the user, ensuring that the educational information not only matches the user's current knowledge level but also stimulates their learning motivation, effectively promoting user participation on the financial education platform.

[0054] Optionally, in the method for generating financial knowledge information provided in Embodiment 1 of this application, before inputting the first behavioral feature into the target evaluation model and outputting the target score, the method further includes: preprocessing historical learning behavior data of historical users to obtain historical behavioral features, and constructing a training set based on the historical behavioral features and historical target scores; encrypting the training set, training a local preset model using the encrypted training set, encrypting the trained model parameters to obtain encrypted parameters; uploading the encrypted parameters to the server and receiving fusion parameters from the server, wherein the server decrypts the encrypted parameters uploaded by each client and fuses the decrypted model parameters to obtain fusion parameters; updating the model parameters of the preset model based on the fusion parameters, training the updated preset model using the encrypted training set, uploading the encrypted model parameters back to the server, iteratively training the preset model based on the fusion parameters from the server until the model converges to obtain the target evaluation model.

[0055] In this embodiment 1, the aim is to establish a financial education goal assessment model that can both protect user privacy and provide personalized assessment. First, the historical user learning behavior data accumulated on the platform is deeply preprocessed. This process includes cleaning, standardization, and screening to ensure data quality. From this data, historical behavioral features that reflect users' learning habits and their level of financial knowledge are extracted. Subsequently, based on these historical behavioral features and known historical target scores, a detailed training dataset is constructed as the basis for model training.

[0056] Then, homomorphic encryption is applied to the constructed training set. This is a technique that allows data processing in an encrypted state. Using this encrypted training set as input, the pre-defined model in the local environment begins the learning process. The parameters produced by the model training are also encrypted, forming encrypted parameters. This is a security guarantee for data transmission in the federated learning mechanism.

[0057] Next, the client securely uploads the encrypted parameters to the central server, while simultaneously receiving the fusion parameters from the server. The server then uses an aggregation algorithm to fuse all the encrypted parameters uploaded by the clients, generating fusion parameters that integrate the learning outcomes of multiple clients, and sends these fusion parameters back to each client.

[0058] Finally, after receiving the fused parameters, the client updates the parameter set of the preset model accordingly and trains it again using the encrypted training set. The trained model is then uploaded to the server again, and the server fuses the model parameters and sends out the fused parameters again. This process is repeated until the model reaches the predetermined convergence criterion, that is, the output results tend to stabilize and there is no longer any significant improvement. At this point, the model is considered to have completed training, and the target evaluation model is obtained.

[0059] For example, the training process of the target evaluation model takes federated learning as an example. The server initializes the global model parameters θ(0). The global parameters θ refer to the set of model weights and biases stored on the server. It is not a specific "value", but a "complete set of all internal parameters of the model". Each user i uses its own local data. Training the model and updating local parameters can be represented as follows: Where η is the learning rate. It is a loss function (such as cross-entropy, mean squared error). This represents the parameters of the global model at the previous time step t-1. Indicates parameters Regarding local datasets loss function The gradient.

[0060] Differential privacy does not directly modify the input data, but instead introduces noise into the uploaded gradient / parameter updates. For example, when the user computes the parameter gradient locally... Then, before uploading, the following will be executed: ,in, This represents the local model parameters for user i at time t. This represents a value with a mean of 0 and a variance of . Gaussian random noise (normal distribution).

[0061] Homomorphic encryption is performed locally on the user's local machine, upon inputting data. It will be encrypted first, which can be represented as: , This represents the encryption algorithm. During model training, computation is performed directly in the ciphertext space, for example: The final result is obtained by decryption by the server. The data remains encrypted throughout the entire transmission and computation process.

[0062] Then, the server performs a weighted average of the parameter updates uploaded by all client users. The updated global parameters are then distributed to each client, and the above steps are repeated until the model converges, resulting in the target evaluation model described above.

[0063] Through the above steps, the technical effect of efficiently integrating multi-source data for model training and updates while protecting user privacy is achieved. This objective evaluation model can not only accurately capture users' learning behavior characteristics, but also quantify users' mastery of financial knowledge and their preferences for reward timeliness, providing decision support for the generation and dynamic adjustment of subsequent personalized educational content. The entire process strictly adheres to data encryption and privacy protection principles, ensuring the security of user data. At the same time, the use of the federated learning mechanism to achieve distributed training of the model improves the model's accuracy and generalization ability, bringing a more intelligent and personalized educational strategy to the financial education platform.

[0064] Optionally, in the method for generating financial knowledge information provided in Embodiment 1 of this application, extracting a second behavioral feature from user behavior data includes: extracting a first feature vector and a second feature vector from user behavior data through a feature extraction model, wherein the first feature vector represents the target user's answer error situation, and the second feature vector represents the target user's response behavior information to answer errors; adjusting the initial behavioral coefficient based on the first behavioral feature to obtain a target behavioral coefficient; and generating a second behavioral feature based on the first feature vector, the second feature vector, and the target behavioral coefficient.

[0065] In this embodiment 1, the aim is to refine the analysis of users' learning performance, especially their understanding and response to errors, in order to more accurately adjust educational strategies. First, a pre-trained feature extraction model is used to deeply analyze user behavior data. This model can identify different types of learning behavior patterns contained in the data, obtaining a first feature vector and a second feature vector. The first feature vector reflects the user's error rate and error type when answering questions at the learning behavior level. If users frequently fail while learning on the platform, they will experience negative emotions. For example, if user A studies for 10 minutes, answers 2 questions correctly and 8 incorrectly, they will resist the learning behavior; if user B studies for 10 minutes, answers 8 questions correctly and 2 incorrectly, they will be more willing to continue participating in the learning behavior. The second feature vector focuses on the user's specific reactions to errors, including the frequency of review after errors, remedial measures, and emotional fluctuations. This vector reveals the user's learning attitude and self-correction mechanism.

[0066] Then, adjust the initial behavior coefficients based on the first behavior feature. It can also generate behavior coefficients for each user. Behavioral coefficients are parameters used to quantify user behavioral tendencies and learning efficiency. The adjustment process considers the frequency, timing, and type of user errors to ensure the model more accurately reflects the user's true learning state and potential needs, thus obtaining target behavioral coefficients that more closely reflect the user's actual learning situation. For example, >1 indicates that user i is particularly resistant to learning behavior after answering a question incorrectly, and therefore engages in learning behavior less frequently. ≈1 indicates that user i generally resists learning behavior after answering a question incorrectly, and the number of times they engage in learning behavior is moderate. <1 indicates that user i is less resistant to learning behavior after answering a question incorrectly, and the number of times they engage in learning behavior actually increases.

[0067] Secondly, the first feature vector, the second feature vector, and the target behavior coefficient are integrated to generate a second behavioral feature that comprehensively describes the user's error response behavior. This process involves multi-level information fusion, which not only analyzes the user's weaknesses in knowledge acquisition but also examines their coping strategies and psychological state when facing learning challenges, making the second behavioral feature a comprehensive indicator that covers the entire dynamics of the user's learning.

[0068] Finally, through the above steps, the platform can accurately capture users' learning difficulties and coping patterns, achieving a deeper technical understanding of user behavior. This in-depth understanding not only helps diagnose users' learning obstacles but also guides the optimization of content recommendation systems, ensuring that educational information targets knowledge gaps while encouraging users to overcome psychological barriers and make continuous progress. Simultaneously, by dynamically adjusting behavioral coefficients, the model can adaptively follow changes in the user's learning trajectory, providing more personalized guidance and support. This significantly improves the effectiveness of educational interventions and user experience, enabling precise customization and efficient implementation of educational strategies.

[0069] Optionally, in the method for generating financial knowledge information provided in Embodiment 1 of this application, generating financial knowledge information to be learned in a financial education platform based on a first behavioral feature, a second behavioral feature, and a target score includes: determining learning effectiveness indicators based on user behavior data, wherein the learning effectiveness indicators include at least a first indicator, a second indicator, and a third indicator, where the first indicator represents the target user's correct answer rate, the second indicator represents the target user's understanding of financial knowledge, and the third indicator represents the target user's course completion level in the financial education platform; determining a candidate knowledge set based on the financial knowledge information in the financial education platform; calculating the score distribution of candidate knowledge information in the candidate knowledge set based on the candidate knowledge set, the learning effectiveness indicators, the first behavioral feature, the second behavioral feature, and the target score; and filtering the candidate knowledge information in the candidate knowledge set based on the score distribution to obtain the financial knowledge information to be learned.

[0070] In this Example 1, the aim is to develop the most suitable learning content for users on the financial education platform to promote the deepening of their financial knowledge and the improvement of their skills. First, user behavior data is analyzed to determine learning effectiveness indicators. This indicator system includes three key dimensions: the first indicator reflects the accuracy of users' answers, or the correctness rate; the second indicator considers the degree to which users have mastered financial knowledge, or the degree of mastery of knowledge points; and the third indicator focuses on the user's learning persistence and completion rate, assessing the user's learning progress and participation through statistical course completion rate, or course completion rate.

[0071] Then, based on the platform's financial knowledge information, a candidate knowledge set is constructed. This set summarizes all learning content that can be recommended to users, including courses of different difficulty levels, video explanations of key knowledge points, and related practice questions. The establishment of the candidate knowledge set aims to cover a wide range of learning fields and diverse learning needs, ensuring that the system has sufficient options to meet users' personalized content requirements.

[0072] Secondly, the score distribution for each candidate knowledge item is calculated based on the candidate knowledge set, learning effectiveness indicators, the user's first behavioral characteristic (learning behavior characteristics), second behavioral characteristic (answer error feedback), and target score. This process involves matching and evaluating the knowledge content with the user's learning status. By analyzing the user's learning performance, knowledge mastery, and error handling methods, combined with their preferences for reward timeliness and depth of financial knowledge, each candidate knowledge item is scored, resulting in a detailed score distribution.

[0073] Finally, based on the score distribution of the candidate knowledge information, the knowledge content with the highest score is selected from the candidate knowledge set and identified as the financial knowledge information to be learned. This selection process ensures that the content recommended to users not only matches their current learning level and stimulates their learning interest, but also takes into account their learning habits and error coping strategies, thereby improving the personalization of the content and learning efficiency.

[0074] Through the above steps, the technology effectively recommends financial knowledge information tailored to users' learning status and preferences. Precise indicator evaluation and knowledge filtering ensure the relevance and effectiveness of educational content, thereby optimizing the user's learning path, increasing the frequency of user interaction with the financial education platform, and, through federated learning algorithms, completing the entire process while protecting user privacy, enhancing user data security, and comprehensively improving the service quality and user experience of the financial education platform.

[0075] Optionally, in the method for generating financial knowledge information provided in Embodiment 1 of this application, calculating the score distribution of candidate knowledge information in the candidate knowledge set based on the candidate knowledge set, learning effect index, first behavioral feature, second behavioral feature, and target score includes: determining a first score based on the candidate knowledge information, learning effect index, and a first function, wherein the first function is used to calculate the degree of matching between the learning difficulty of the candidate knowledge information and the knowledge already mastered by the target user; determining a second score based on the candidate knowledge information, learning behavioral feature, and second function, wherein the learning behavioral feature includes at least one of the following: target score, second behavioral feature, and the second function is used to calculate the degree of matching between the candidate knowledge information and the learning behavioral feature; determining a third score based on the candidate knowledge information, first behavioral feature, and third function, wherein the third function is used to calculate the degree of interest of the target user in the candidate knowledge information; and calculating the score distribution of the candidate knowledge information based on the first score, second score, and third score.

[0076] In this embodiment 1, in order to accurately match the user's learning needs with the course content of the financial education platform and ensure the appropriateness and attractiveness of the recommended knowledge, firstly, based on candidate knowledge information and learning effect indicators, a first function is applied. The first score is generated by calculating the match between the learning difficulty and the user's knowledge level. This score quantifies whether the difficulty of the knowledge content is appropriate for the user's current level of understanding, aiming to avoid the frustration caused by excessive difficulty or the boredom caused by excessive difficulty, ensuring that the content can stimulate the user's learning interest and effectively promote the deepening of knowledge.

[0077] Then, the degree of fit between the candidate knowledge information and the user's specific learning behavior pattern is calculated through the second function. Calculate the second score. This score is determined based on the user's preference for reward timeliness, reaction to errors, and efficiency of knowledge acquisition. It delves into the user's learning motivation and behavioral habits, ensuring that the recommended content is not only appropriate at the knowledge level but also highly relevant to the user at the behavioral incentive level, thereby enhancing learning engagement and satisfaction.

[0078] Secondly, based on candidate knowledge information, first behavioral characteristics (users' learning habits and preferences), and third functions... A third score is determined, reflecting the user's level of interest in specific knowledge content. By deeply analyzing the user's historical behavior, the system can identify the user's preferred topics, learning styles, and interaction patterns, thereby generating the third score. This ensures that recommended course content can capture the user's attention and improve learning initiative and persistence.

[0079] Finally, the score distribution of the candidate knowledge information is calculated by combining the first, second, and third scores. For example, the calculation process can be represented as follows: ,in, Candidate knowledge set , This represents the feature vector of user u's learning behavior. This represents a metric indicating the learning performance of user u. , This represents the second eigenvector mentioned above. This represents the first fraction mentioned above. This refers to the second fraction mentioned above. Represents the weight parameters. This means selecting, from all available candidate knowledge, the content that will allow function S to achieve the highest score. This refers to the aforementioned financial knowledge information to be learned. This distribution integrates multiple dimensions of consideration, including knowledge difficulty, behavioral incentives, and user interests, providing a comprehensive basis for content recommendation decisions. By comparing the score distributions of different knowledge information, the system can select the course combinations that best meet the user's personalized needs, achieving precise delivery of educational content.

[0080] Through the above steps, the technology achieves the goal of intelligently recommending educational content to users on the financial education platform that is appropriate to their knowledge level, effectively stimulates their learning motivation, and highly aligns with their personal interests and preferences. This achievement is based not only on a deep understanding of users' dynamic learning behavior but also on the precise matching of knowledge difficulty and learning patterns, ensuring the personalization and effectiveness of the educational content. This optimizes the user's learning experience, increases user engagement on the financial education platform, and, moreover, protects user privacy throughout the entire recommendation process, demonstrating the human-centered and secure nature of the technology. Ultimately, this comprehensively enhances the service quality and user satisfaction of the financial education platform.

[0081] Optionally, in the method for generating financial knowledge information provided in Embodiment 1 of this application, after generating the financial knowledge information to be learned in the financial education platform based on the first behavioral feature, the second behavioral feature, and the target score, the method further includes: determining the number of target knowledge points already learned by the target user based on user behavior data, and determining the course completion progress based on the number of target knowledge points; determining the score information of each knowledge point based on user behavior data, and calculating the average score based on the score information of each knowledge point; determining the learning time of the target user based on user behavior data, and determining the level of enthusiasm of the target user's learning behavior; determining the display method of the financial knowledge information to be learned based on the course completion progress, average score, learning time, enthusiasm, second behavioral feature, and target score, and pushing the financial knowledge information to be learned to the target user based on the display method.

[0082] In this embodiment 1, in order to optimize the presentation of financial knowledge information, firstly, by analyzing user behavior data, the number of target knowledge points that the target user has learned is determined, thereby calculating the course completion progress. For example, the course completion progress calculation process can be expressed as: ,in, Indicates the number of target knowledge points. This represents the total number of knowledge points. Further analysis of user behavior data is needed to calculate the score for each knowledge point. And based on this information, the average score is derived. The calculation process can be expressed as follows: The average score reflects a user's overall mastery of the learned knowledge points and is a key indicator for assessing a user's knowledge level, providing a quantitative basis for adjusting subsequent presentation methods. Further analysis of user behavior data to determine user learning time can be represented as: ,in, This indicates the preset recommended learning time. This indicates the user's completed learning time. If the recommended duration is exceeded, the value is set to 1 to avoid infinite growth. Let the initial level of enthusiasm be set. The initial level of motivation is adjusted based on the user's learning behavior within the range [0,1] to arrive at the final level of motivation. For example, if a user studies regularly every day, the motivation level increases by 0.1; if a user repeatedly reviews incorrect questions, the motivation level increases by 0.2; if a user frequently skips questions, the motivation level decreases by 0.2; if a user quits halfway through, the motivation level decreases by 0.3.

[0083] Then, by comprehensively considering course completion progress, average score, learning time, the level of engagement in learning behavior, the user's secondary behavioral characteristics, and primary score, the presentation method of the financial knowledge information to be learned is determined. The presentation method may include the order in which knowledge points are presented, the depth of content, interactive formats, and feedback mechanisms. The presentation method is customized based on the user's specific situation, such as whether they need more review of basic concepts or are suited to more advanced case analysis. Through this approach, the most appropriate financial knowledge information is pushed to the target user, ensuring that the content not only meets their current knowledge level but also stimulates their learning interest and improves learning efficiency.

[0084] For example, the overall learning progress can be calculated based on course completion progress, average score, learning time, and level of engagement. The calculation formula can be expressed as follows: Among them, weight , , , This can be determined through regression or machine learning using historical learning data. When If the value is less than 0.4, basic financial knowledge and high-reward feedback content can be pushed; if 0.4 ≤ If the score is less than 0.7, then intermediate case studies and moderately challenging content can be pushed; when... If the score is ≥0.7, then comprehensive applications and open-ended tasks can be pushed. When the second behavioral feature λ is large, a reward-driven presentation method can be adopted, such as giving points / visual rewards after learning; when the first score β is small, an instant feedback method can be adopted, such as explaining wrong questions immediately; when the second score γ is large, error correction reminders and challenging problems can be added.

[0085] Through the above steps, the technology effectively adjusts the way knowledge information is displayed based on the user's learning status and behavioral characteristics. A deep understanding of users' learning habits and knowledge mastery ensures personalized presentation and efficient absorption of educational content, significantly improving the user's learning experience and the platform's educational effectiveness. Simultaneously, the entire process is conducted while protecting user data privacy, demonstrating the security and compliance of the technology, enhancing user trust in the platform, and laying a solid foundation for improving the quality of financial education and popularizing financial knowledge.

[0086] Optionally, in the method for generating financial knowledge information provided in Embodiment 1 of this application, after pushing the financial knowledge information to be learned to the target user according to the display method, the method further includes: after detecting that the target user has completed the learning process of the financial knowledge information to be learned, collecting the target user's transaction information; inputting the target user's transaction information and the financial knowledge information already learned by the target user into the target recommendation model, and outputting the financial knowledge recommendation information, wherein the target recommendation model is a model trained on a preset model using historical users' transaction information and historical users' learned financial knowledge information based on a federated learning algorithm.

[0087] In this embodiment 1, to improve the accuracy of educational content recommendations on the financial education platform and promote the integration of users' financial knowledge with practical skills, firstly, after a user completes the learning of the financial knowledge to be learned, the user's recent transaction information is collected. This information includes the user's transaction records on the financial platform, such as transaction frequency, transaction amount, investment preferences, and risk tolerance.

[0088] Then, the acquired user transaction information and their learned financial knowledge are input into the target recommendation model. The target recommendation model is trained based on a federated learning algorithm, integrating historical user data, including transaction information and learning history. This model can efficiently learn using distributed data while protecting user privacy, in order to generate more personalized financial knowledge recommendations.

[0089] Secondly, the target recommendation model uses the input information to perform in-depth analysis and prediction, outputting financial knowledge recommendations tailored to the user. This recommendation generation process comprehensively considers the user's learning outcomes, trading habits, and market dynamics, ensuring that the recommended content not only matches the user's financial knowledge level but also guides their decision-making in practice.

[0090] Finally, based on the output of the target recommendation model, personalized financial knowledge recommendations are pushed to users. These recommendations may include more in-depth theoretical explanations, analysis of specific financial products, interpretation of market trends, or guidance on investment strategies, aiming to help users apply what they have learned to practice and improve their financial decision-making capabilities.

[0091] Through the above steps, the technology achieves the goal of intelligently generating personalized financial knowledge recommendations based on users' learning outcomes and transaction behavior. This not only promotes the close integration of theoretical knowledge and practical operation but also demonstrates that federated learning algorithms, while protecting user privacy, can effectively utilize the advantages of distributed data to achieve intelligent and personalized educational content recommendations. Simultaneously, the entire recommendation process adheres to data security and privacy protection principles, providing users with a safe and efficient learning environment, further enhancing the platform's user satisfaction and market competitiveness.

[0092] Optionally, in the method for generating financial knowledge information provided in Embodiment 1 of this application, user behavior data of the target user on the financial education platform is collected, and feature extraction is performed on the user behavior data to obtain the first behavioral feature. This includes: obtaining user information of the target user, wherein the user information includes at least one of the following: age information, educational background information, occupation, and financial knowledge level, wherein the financial knowledge level is determined based on a financial knowledge test questionnaire; collecting user behavior data generated by the target user on various clients of the financial education platform, wherein the user behavior data includes at least one of the following: learning duration, learning completion rate, click data, learning feedback data, and transaction behavior information; performing data preprocessing operations on the user behavior data, and performing feature extraction operations on the processed user behavior data to obtain the first behavioral feature.

[0093] In this embodiment 1, in order to construct an accurate user profile and support the generation of personalized financial knowledge education content, the user information of the target user is first retrieved from the database. This information covers age, educational background, occupation, and financial knowledge level assessed through a financial knowledge test questionnaire. The purpose is to fully understand the user's basic attributes and knowledge starting point.

[0094] Then, the interaction records of target users on various clients of the financial education platform are continuously monitored and collected to form a user behavior dataset. These datasets include users' learning time, the proportion of courses completed, their click preferences on educational content, submitted learning feedback, and transaction behavior information on the platform. The aim is to capture users' learning habits, interests, and practical applications, providing rich material for subsequent personalized content recommendations.

[0095] Finally, the collected user behavior data undergoes preprocessing operations, including cleaning, standardization, and missing value imputation, to eliminate data noise and outliers, ensuring data quality and consistency. The preprocessed data is then transformed into primary behavioral features through feature extraction, extracting feature vectors closely related to user learning outcomes from massive datasets, such as learning frequency, knowledge mastery level, and transaction decision-making patterns.

[0096] Through the steps described above, a rich user profile is constructed from two dimensions: basic user information and behavioral data, enabling precise content recommendations. Utilizing advanced technology to mine and leverage user data achieves personalized and intelligent education. This not only provides users with efficient and comfortable learning content but also enhances user activity and engagement.

[0097] Optionally, in this embodiment 1, Figure 3 This is a schematic diagram illustrating the optional process for generating financial knowledge information to be recommended in a financial education platform, according to Embodiment 1 of this application. For example... Figure 3 As shown, the platform collects user learning behavior data on its financial education platform, covering key indicators such as learning time, frequency, and accuracy rate. This data is collected through the platform's logs and user activity traces. Then, using a machine learning model adapted to encrypted data, the platform performs in-depth analysis of user behavior data to identify user learning habits and behavioral patterns, such as the aforementioned first feature vector, second feature vector, and target score. Next, based on the analysis results of user behavior characteristics, the platform customizes educational content. Then, focusing on the application of the federated learning framework, the process of model training based on user distribution data is described, while simultaneously protecting user privacy. Finally, the difficulty of the educational content is adjusted in real time based on user learning progress and feedback. The platform continuously optimizes educational content and system performance through user feedback. By collecting user feedback during platform use and adjusting and optimizing educational content and system performance based on the feedback results, the platform further improves user learning effectiveness and satisfaction.

[0098] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0099] In summary, the method for generating financial knowledge information provided in this application, with the authorization of the target user, collects user behavior data of the target user on a financial education platform, and extracts features from the user behavior data to obtain a first behavioral feature, wherein the first behavioral feature represents the learning behavior characteristics of the target user on the financial education platform; the first behavioral feature is input into a target evaluation model to output a target score, wherein the target score includes at least a first score and a second score, wherein the first score represents the target user's preference information for reward timeliness, and the second score represents the target user's mastery of financial knowledge, and the target evaluation model is obtained by training a preset model based on a federated learning algorithm according to the historical user learning behavior of the financial education platform; a second behavioral feature is extracted from the user behavior data, wherein the second behavioral feature represents the characteristics of the target user's response behavior to incorrect answers; and financial knowledge information to be learned in the financial education platform is generated based on the first behavioral feature, the second behavioral feature, and the target score. This solves the problem in related technologies that the learning content in financial education platforms is usually uniformly designed, lacking consideration for different user needs and different learning behaviors, resulting in low user participation in financial education platforms.

[0100] By inputting the first behavioral feature into a target evaluation model trained by a federated learning algorithm, a target score encompassing the user's reward timeliness preference and financial knowledge mastery can be output, achieving personalized educational needs assessment and further optimizing the allocation of educational resources. Furthermore, by extracting features from the user's response behavior to incorrect answers, a second behavioral feature is formed, providing direct evidence for identifying the user's learning difficulties and improvement directions. The financial education platform's learning content, generated based on the first and second behavioral features and the target score, can highly match the user's learning ability and preferences, achieving personalized recommendations of educational content, improving user learning efficiency and satisfaction, and ultimately increasing user participation in the financial education platform.

[0101] Example 2

[0102] This application also provides a financial knowledge information generation apparatus. It should be noted that the financial knowledge information generation apparatus of this application can be used to execute the financial knowledge information generation method provided in this application. The following describes the financial knowledge information generation apparatus provided in this application.

[0103] According to an embodiment of this application, an apparatus for implementing the above-described method for generating financial knowledge information is also provided. Figure 4 This is a schematic diagram of a financial knowledge information generation device provided according to Embodiment 2 of this application. Figure 4As shown, the device includes: a first acquisition unit 401, an evaluation unit 402, an extraction unit 403, and a generation unit 404.

[0104] Specifically, the first collection unit 401 is used to collect user behavior data of the target user on the financial education platform when the target user is authorized, and to extract features from the user behavior data to obtain the first behavior feature, wherein the first behavior feature represents the learning behavior feature of the target user on the financial education platform.

[0105] Evaluation unit 402 is used to input the first behavioral feature into the target evaluation model and output the target score. The target score includes at least a first score and a second score. The first score represents the target user's preference information for reward timeliness, and the second score represents the target user's mastery of financial knowledge. The target evaluation model is obtained by training a preset model based on the historical user learning behavior of the financial education platform using a federated learning algorithm.

[0106] Extraction unit 403 is used to extract a second behavioral feature from user behavior data, wherein the second behavioral feature represents the characteristics of the target user's response behavior to incorrect answers.

[0107] The generation unit 404 is used to generate financial knowledge information to be learned in the financial education platform based on the first behavioral feature, the second behavioral feature, and the target score.

[0108] The financial knowledge information generation device provided in this application embodiment collects user behavior data of the target user on a financial education platform with the authorization of the target user by a first acquisition unit 401, and extracts features from the user behavior data to obtain a first behavioral feature, wherein the first behavioral feature represents the learning behavior characteristics of the target user on the financial education platform; the evaluation unit 402 inputs the first behavioral feature into a target evaluation model and outputs a target score, wherein the target score includes at least a first score and a second score, wherein the first score represents the target user's preference information for reward timeliness, and the second score represents the target user's mastery of financial knowledge, and the target evaluation model is obtained by training a preset model based on a federated learning algorithm according to the historical user learning behavior of the financial education platform; the extraction unit 403 extracts a second behavioral feature from the user behavior data, wherein the second behavioral feature represents the characteristics of the target user's response behavior to incorrect answers; the generation unit 404 generates financial knowledge information to be learned in the financial education platform based on the first behavioral feature, the second behavioral feature and the target score, thereby solving the problem that the learning content in the financial education platform is usually uniformly designed in related technologies, lacking consideration for different user needs and different learning behaviors, resulting in low user participation in the financial education platform.

[0109] By inputting the first behavioral feature into a target evaluation model trained by a federated learning algorithm, a target score encompassing the user's reward timeliness preference and financial knowledge mastery can be output, achieving personalized educational needs assessment and further optimizing the allocation of educational resources. Furthermore, by extracting features from the user's response behavior to incorrect answers, a second behavioral feature is formed, providing direct evidence for identifying the user's learning difficulties and improvement directions. The financial education platform's learning content, generated based on the first and second behavioral features and the target score, can highly match the user's learning ability and preferences, achieving personalized recommendations of educational content, improving user learning efficiency and satisfaction, and ultimately increasing user participation in the financial education platform.

[0110] Optionally, in the financial knowledge information generation apparatus provided in Embodiment 2 of this application, the apparatus further includes: a construction unit, used to preprocess historical learning behavior data of historical users to obtain historical behavior features before inputting the first behavioral features into the target evaluation model and outputting the target score, and to construct a training set based on the historical behavioral features and historical target scores; a first training unit, used to encrypt the training set, train a local preset model using the encrypted training set, and encrypt the trained model parameters to obtain encrypted parameters; a receiving unit, used to upload the encrypted parameters to the server and receive the fusion parameters sent by the server, wherein the server decrypts the encrypted parameters uploaded by each client and fuses the decrypted model parameters to obtain fusion parameters; a second training unit, used to update the model parameters of the preset model according to the fusion parameters, train the updated preset model using the encrypted training set, upload the encrypted model parameters to the server again, and iteratively train the preset model based on the fusion parameters sent by the server until the model converges to obtain the target evaluation model.

[0111] Optionally, in the financial knowledge information generation apparatus provided in Embodiment 2 of this application, the extraction unit 403 includes: an extraction subunit, used to extract a first feature vector and a second feature vector from user behavior data through a feature extraction model, wherein the first feature vector represents the target user's answer error and the second feature vector represents the target user's response behavior information in response to the answer error; an adjustment subunit, used to adjust the initial behavior coefficient according to the first behavior feature to obtain a target behavior coefficient; and a generation subunit, used to generate a second behavior feature based on the first feature vector, the second feature vector and the target behavior coefficient.

[0112] Optionally, in the financial knowledge information generation device provided in Embodiment 2 of this application, the generation unit 404 includes: a first determining subunit, used to determine learning effect indicators based on user behavior data, wherein the learning effect indicators include at least: a first indicator, a second indicator, and a third indicator, wherein the first indicator represents the target user's correct answer rate, the second indicator represents the target user's understanding of financial knowledge, and the third indicator represents the target user's course completion level in the financial education platform; a second determining subunit, used to determine a candidate knowledge set based on the financial knowledge information of the financial education platform; a calculation subunit, used to calculate the score distribution of candidate knowledge information in the candidate knowledge set based on the candidate knowledge set, the learning effect indicators, the first behavioral feature, the second behavioral feature, and the target score; and a filtering subunit, used to filter the candidate knowledge information in the candidate knowledge set based on the score distribution of the candidate knowledge information to obtain the financial knowledge information to be learned.

[0113] Optionally, in the financial knowledge information generation apparatus provided in Embodiment 2 of this application, the aforementioned calculation subunit includes: a first determining module, used to determine a first score based on candidate knowledge information, learning effect indicators, and a first function, wherein the first function is used to calculate the degree of matching between the learning difficulty of the candidate knowledge information and the knowledge already mastered by the target user; a second determining module, used to determine a second score based on candidate knowledge information, learning behavior characteristics, and a second function, wherein the learning behavior characteristics include at least one of the following: target score, second behavior characteristics, and the second function is used to calculate the degree of matching between the candidate knowledge information and the learning behavior characteristics; a third determining module, used to determine a third score based on candidate knowledge information, the first behavior characteristics, and a third function, wherein the third function is used to calculate the degree of interest of the target user in the candidate knowledge information; and a calculation module, used to calculate the score distribution of the candidate knowledge information based on the first score, the second score, and the third score.

[0114] Optionally, in the financial knowledge information generation apparatus provided in Embodiment 2 of this application, the apparatus further includes: a first determining unit, configured to, after generating financial knowledge information to be learned in the financial education platform based on the first behavioral characteristics, the second behavioral characteristics, and the target score, determine the number of target knowledge points already learned by the target user based on user behavior data, and determine the course completion progress based on the number of target knowledge points; a calculation unit, configured to determine the score information of each knowledge point based on user behavior data, and calculate the average score based on the score information of each knowledge point; a second determining unit, configured to determine the learning time of the target user based on user behavior data, and determine the level of enthusiasm of the target user's learning behavior; and a push unit, configured to determine the display method of the financial knowledge information to be learned based on the course completion progress, average score, learning time, enthusiasm, the second behavioral characteristics, and the target score, and push the financial knowledge information to be learned to the target user based on the display method.

[0115] Optionally, in the financial knowledge information generation device provided in Embodiment 2 of this application, the device further includes: a second acquisition unit, used to acquire the target user's transaction information after the target user has completed the learning process of the financial knowledge information to be learned after the target user pushes the financial knowledge information to be learned to the target user according to the display method; and a processing unit, used to input the target user's transaction information and the financial knowledge information already learned by the target user into a target recommendation model and output financial knowledge recommendation information, wherein the target recommendation model is a model trained on a preset model using historical users' transaction information and historical users' financial knowledge information already learned by the federated learning algorithm.

[0116] Optionally, in the financial knowledge information generation device provided in Embodiment 2 of this application, the first collection unit 401 mentioned above includes: an acquisition unit, used to acquire user information of the target user, wherein the user information includes at least one of the following: age information, educational background information, occupation, and financial knowledge level, wherein the financial knowledge level is determined based on a financial knowledge test questionnaire; a third collection unit, used to collect user behavior data generated by the target user on various clients of the financial education platform, wherein the user behavior data includes at least one of the following: learning duration, learning completion rate, click data, learning feedback data, and transaction behavior information; and a processing unit, used to perform data preprocessing operations on the user behavior data and perform feature extraction operations on the processed user behavior data to obtain a first behavior feature.

[0117] It should be noted that the first acquisition unit 401, evaluation unit 402, extraction unit 403, and generation unit 404 mentioned above correspond to steps S201 to S204 in Embodiment 1. The two modules and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above modules can also be part of the device and run in the computer terminal 10 provided in Embodiment 1.

[0118] Example 3

[0119] Embodiments of this application may provide an electronic device. Figure 5 This is a schematic diagram of an electronic device for generating financial knowledge information according to Embodiment 3 of this application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5 (Only one is shown) processor 502, memory 504, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0120] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0121] The processor can access information and applications stored in the memory via a transmission device to execute the following steps: With the authorization of the target user, collect user behavior data of the target user on the financial education platform, and extract features from the user behavior data to obtain a first behavioral feature, wherein the first behavioral feature represents the target user's learning behavior characteristics on the financial education platform; input the first behavioral feature into the target evaluation model and output a target score, wherein the target score includes at least a first score and a second score, wherein the first score represents the target user's preference information for reward timeliness, and the second score represents the target user's mastery of financial knowledge; the target evaluation model is obtained by training a preset model based on a federated learning algorithm according to the historical user learning behavior on the financial education platform; extract a second behavioral feature from the user behavior data, wherein the second behavioral feature represents the characteristics of the target user's response behavior to incorrect answers; generate financial knowledge information to be learned in the financial education platform based on the first behavioral feature, the second behavioral feature, and the target score.

[0122] The processor can access information and applications stored in the memory via a transmission device to execute the following steps: Before inputting the first behavioral feature into the target evaluation model and outputting the target score, the method further includes: preprocessing historical learning behavior data of historical users to obtain historical behavioral features, and constructing a training set based on the historical behavioral features and historical target scores; encrypting the training set, training a local preset model using the encrypted training set, encrypting the trained model parameters to obtain encrypted parameters; uploading the encrypted parameters to the server and receiving fusion parameters from the server, wherein the server decrypts the encrypted parameters uploaded by each client and fuses the decrypted model parameters to obtain fusion parameters; updating the model parameters of the preset model based on the fusion parameters, training the updated preset model using the encrypted training set, uploading the encrypted model parameters back to the server, iteratively training the preset model based on the fusion parameters from the server until the model converges to obtain the target evaluation model.

[0123] The processor can access information and applications stored in the memory via a transmission device to perform the following steps: extracting a second behavioral feature from user behavior data, including: extracting a first feature vector and a second feature vector from the user behavior data using a feature extraction model, wherein the first feature vector represents the target user's answer errors and the second feature vector represents the target user's response behavior information to answer errors; adjusting the initial behavioral coefficients based on the first behavioral feature to obtain target behavioral coefficients; and generating the second behavioral feature based on the first feature vector, the second feature vector, and the target behavioral coefficient.

[0124] The processor can invoke information and applications stored in the memory via a transmission device to execute the following steps: generating financial knowledge information to be learned in the financial education platform based on a first behavioral feature, a second behavioral feature, and a target score, including: determining learning effectiveness indicators based on user behavior data, wherein the learning effectiveness indicators include at least: a first indicator, a second indicator, and a third indicator, where the first indicator represents the target user's correct answer rate, the second indicator represents the target user's understanding of financial knowledge, and the third indicator represents the target user's course completion level in the financial education platform; determining a candidate knowledge set based on the financial knowledge information in the financial education platform; calculating the score distribution of candidate knowledge information in the candidate knowledge set based on the candidate knowledge set, the learning effectiveness indicators, the first behavioral feature, the second behavioral feature, and the target score; and filtering the candidate knowledge information in the candidate knowledge set based on the score distribution to obtain the financial knowledge information to be learned.

[0125] The processor can invoke information and applications stored in memory via a transmission device to perform the following steps: calculating the score distribution of candidate knowledge information in the candidate knowledge set based on a candidate knowledge set, a learning effect index, a first behavioral feature, a second behavioral feature, and a target score, including: determining a first score based on candidate knowledge information, a learning effect index, and a first function, wherein the first function is used to calculate the degree of matching between the learning difficulty of the candidate knowledge information and the knowledge already mastered by the target user; determining a second score based on candidate knowledge information, a learning behavioral feature, and a second function, wherein the learning behavioral feature includes at least one of the following: a target score, a second behavioral feature, and the second function is used to calculate the degree of matching between the candidate knowledge information and the learning behavioral feature; determining a third score based on candidate knowledge information, a first behavioral feature, and a third function, wherein the third function is used to calculate the degree of interest of the target user in the candidate knowledge information; and calculating the score distribution of the candidate knowledge information based on the first score, the second score, and the third score.

[0126] The processor can invoke information and applications stored in the memory via a transmission device to perform the following steps: After generating financial knowledge information to be learned in the financial education platform based on the first behavioral characteristics, the second behavioral characteristics, and the target score, the above method further includes: determining the number of target knowledge points already learned by the target user based on user behavior data, and determining the course completion progress based on the number of target knowledge points; determining the score information for each knowledge point based on user behavior data, and calculating the average score based on the score information for each knowledge point; determining the learning time of the target user based on user behavior data, and determining the level of enthusiasm of the target user's learning behavior; determining the display method of the financial knowledge information to be learned based on the course completion progress, average score, learning time, enthusiasm, second behavioral characteristics, and target score, and pushing the financial knowledge information to be learned to the target user according to the display method.

[0127] The processor can access the information and application programs stored in the memory via a transmission device to perform the following steps: After pushing financial knowledge information to be learned to the target user according to the display method, the above method further includes: after detecting that the target user has completed the learning process of the financial knowledge information to be learned, collecting the target user's transaction information; inputting the target user's transaction information and the financial knowledge information already learned by the target user into the target recommendation model, and outputting financial knowledge recommendation information, wherein the target recommendation model is a model trained on a preset model using historical users' transaction information and historical users' financial knowledge information based on a federated learning algorithm.

[0128] The processor can access information and applications stored in the memory via a transmission device to perform the following steps: collecting user behavior data of the target user on the financial education platform and extracting features from the user behavior data to obtain the first behavioral feature, including: obtaining user information of the target user, wherein the user information includes at least one of the following: age information, educational background information, occupation, and financial knowledge level, wherein the financial knowledge level is determined based on a financial knowledge test questionnaire; collecting user behavior data generated by the target user on various clients of the financial education platform, wherein the user behavior data includes at least one of the following: learning duration, learning completion rate, click data, learning feedback data, and transaction behavior information; performing data preprocessing operations on the user behavior data and performing feature extraction operations on the processed user behavior data to obtain the first behavioral feature.

[0129] By employing the embodiments of this application, inputting the first behavioral feature into a target evaluation model trained by a federated learning algorithm, a target score encompassing the user's reward timeliness preference and financial knowledge mastery can be output, achieving personalized educational needs assessment and further optimizing the allocation of educational resources. Furthermore, by extracting features from the user's response behavior regarding incorrect answers, a second behavioral feature is formed, providing direct evidence for identifying the user's learning difficulties and improvement directions. The financial education platform's learning knowledge information generated based on the first and second behavioral features and the target score can highly match the user's learning ability and preferences, achieving personalized recommendations of educational content, improving user learning efficiency and satisfaction, and ultimately increasing user participation in the financial education platform.

[0130] Those skilled in the art will understand that Figure 5 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 5 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 5 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 5 The different configurations shown.

[0131] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0132] Example 4

[0133] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the financial knowledge information generation method provided in Embodiment 1.

[0134] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0135] This application also provides a computer program product, which, when executed on a data processing device, is suitable for performing the steps of a method for generating financial knowledge information.

[0136] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0137] In the above embodiments of this application, the descriptions of each embodiment have their own emphasis. Parts not described in detail in a certain embodiment can be referred to in the relevant descriptions of other embodiments. It should be understood that the disclosed technical content can be implemented in other ways in the several embodiments provided in this application. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection between units or modules, and may be electrical or other forms. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs. In addition, the functional units in the various embodiments of this application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The integrated units described above can be implemented in either hardware or software functional units.

[0138] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0139] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for generating financial knowledge information, characterized in that, include: With the authorization of the target user, user behavior data of the target user on the financial education platform is collected, and feature extraction is performed on the user behavior data to obtain a first behavioral feature, wherein the first behavioral feature represents the learning behavior characteristics of the target user on the financial education platform. The first behavioral feature is input into the target evaluation model, and a target score is output. The target score includes at least a first score and a second score. The first score represents the target user's preference information for reward timeliness, and the second score represents the target user's mastery of financial knowledge. The target evaluation model is obtained by training a preset model based on a federated learning algorithm according to the historical user learning behavior of the financial education platform. Extract a second behavioral feature from the user behavior data, wherein the second behavioral feature represents the characteristics of the target user's response behavior to incorrect answers; Based on the first behavioral feature, the second behavioral feature, and the target score, financial knowledge information to be learned in the financial education platform is generated.

2. The method according to claim 1, characterized in that, Before inputting the first behavioral feature into the target evaluation model and outputting the target score, the method further includes: The historical learning behavior data of historical users are preprocessed to obtain historical behavior features, and a training set is constructed based on the historical behavior features and historical target scores; The training set is encrypted, and the local preset model is trained using the encrypted training set. The parameters of the trained model are then encrypted to obtain encrypted parameters. The encrypted parameters are uploaded to the server, and the server receives the fusion parameters sent by the server. The server decrypts the encrypted parameters uploaded by each client and fuses the decrypted model parameters to obtain the fusion parameters. The model parameters of the preset model are updated according to the fusion parameters. The updated preset model is trained using the encrypted training set. The encrypted model parameters are uploaded to the server again. The preset model is iteratively trained based on the fusion parameters issued by the server until the model converges, and the target evaluation model is obtained.

3. The method according to claim 1, characterized in that, Extracting a second behavioral feature from the user behavior data includes: A first feature vector and a second feature vector are extracted from the user behavior data using a feature extraction model. The first feature vector represents the target user's incorrect answer situation, and the second feature vector represents the target user's response behavior information in response to incorrect answers. The initial behavior coefficients are adjusted based on the first behavior characteristic to obtain the target behavior coefficients; The second behavioral feature is generated based on the first feature vector, the second feature vector, and the target behavioral coefficient.

4. The method according to claim 1, characterized in that, Based on the first behavioral feature, the second behavioral feature, and the target score, financial knowledge information to be learned in the financial education platform is generated, including: Learning effectiveness indicators are determined based on the user behavior data, wherein the learning effectiveness indicators include at least: a first indicator, a second indicator, and a third indicator, wherein the first indicator represents the target user's correct answer rate, the second indicator represents the target user's understanding of financial knowledge, and the third indicator represents the target user's course completion rate on the financial education platform; A candidate knowledge set is determined based on the financial knowledge information of the financial education platform. The score distribution of candidate knowledge information in the candidate knowledge set is calculated based on the candidate knowledge set, the learning effect index, the first behavioral feature, the second behavioral feature, and the target score. The financial knowledge information to be learned is obtained by filtering the candidate knowledge set based on the score distribution of the candidate knowledge information.

5. The method according to claim 4, characterized in that, The score distribution of candidate knowledge information in the candidate knowledge set is calculated based on the candidate knowledge set, the learning effect index, the first behavioral feature, the second behavioral feature, and the target score, including: A first score is determined based on the candidate knowledge information, the learning effect index, and a first function, wherein the first function is used to calculate the degree of matching between the learning difficulty of the candidate knowledge information and the knowledge already mastered by the target user; A second score is determined based on the candidate knowledge information, learning behavior features, and a second function, wherein the learning behavior features include at least one of the following: the target score, the second behavior feature, and the second function is used to calculate the degree of matching between the candidate knowledge information and the learning behavior feature; A third score is determined based on the candidate knowledge information, the first behavioral feature, and the third function, wherein the third function is used to calculate the target user's degree of interest in the candidate knowledge information; The score distribution of the candidate knowledge information is calculated based on the first score, the second score, and the third score.

6. The method according to claim 1, characterized in that, After generating the financial knowledge information to be learned in the financial education platform based on the first behavioral feature, the second behavioral feature, and the target score, the method further includes: Based on the user behavior data, determine the number of target knowledge points that the target user has learned, and determine the course completion progress based on the number of target knowledge points; The score information for each knowledge point is determined based on the user behavior data, and the average score is calculated based on the score information for each knowledge point. Based on the user behavior data, determine the learning time of the target user and the level of enthusiasm of the target user's learning behavior; The presentation method of the financial knowledge information to be learned is determined based on the course completion progress, the average score, the learning time, the level of enthusiasm, the second behavioral characteristic, and the target score, and the financial knowledge information to be learned is pushed to the target user according to the presentation method.

7. The method according to claim 6, characterized in that, After pushing the financial knowledge information to be learned to the target user according to the aforementioned display method, the method further includes: After detecting that the target user has completed the learning process of the financial knowledge information to be learned, the transaction information of the target user is collected; The target user's transaction information and the financial knowledge information that the target user has learned are input into the target recommendation model, and the financial knowledge recommendation information is output. The target recommendation model is a model trained on a preset model using historical users' transaction information and the financial knowledge information that historical users have learned, based on a federated learning algorithm.

8. The method according to claim 1, characterized in that, Collect user behavior data of the target user on the financial education platform, and extract features from the user behavior data to obtain the first behavioral features, including: Obtain user information of the target user, wherein the user information includes at least one of the following: age information, educational background information, occupation, and financial knowledge level, wherein the financial knowledge level is determined based on a financial knowledge test questionnaire; Collect user behavior data generated by the target user on various clients of the financial education platform, wherein the user behavior data includes at least one of the following: learning duration, learning completion rate, click data, learning feedback data, and transaction behavior information; The user behavior data is preprocessed, and the processed user behavior data is then subjected to feature extraction to obtain the first behavior feature.

9. A device for generating financial knowledge information, characterized in that, include: The first collection unit is used to collect user behavior data of the target user on the financial education platform when the target user is authorized, and to extract features from the user behavior data to obtain a first behavior feature, wherein the first behavior feature represents the learning behavior characteristics of the target user on the financial education platform. An evaluation unit is used to input the first behavioral feature into a target evaluation model and output a target score. The target score includes at least a first score and a second score. The first score represents the target user's preference information for reward timeliness, and the second score represents the target user's mastery of financial knowledge. The target evaluation model is obtained by training a preset model based on a federated learning algorithm according to the historical user learning behavior of the financial education platform. An extraction unit is configured to extract a second behavioral feature from the user behavior data, wherein the second behavioral feature represents the characteristics of the target user's response behavior to incorrect answers; The generation unit is used to generate financial knowledge information to be learned in the financial education platform based on the first behavioral feature, the second behavioral feature, and the target score.

10. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the method for generating financial knowledge information according to any one of claims 1 to 8.

Images