Information processing system, method and equipment based on bank loan platform and medium

Abstract

The invention relates to the technical field of information processing, and discloses an information processing system and method based on a bank loan platform, equipment and a medium. The system comprises a data burying point design module used for determining a data burying point scheme according to a loan application process and user behavior characteristics; the data acquisition module is used for acquiring user data in real time according to the data burying point scheme and sending the user data to the back-end server; the data processing and analysis module is used for receiving, processing and analyzing the user data to determine a data analysis result; the result application module comprises a loan approval optimization unit which is used for optimizing the loan approval process according to the data analysis result, and the loan approval process comprises the steps of adjusting the loan amount and the interest rate and making a loan approval decision to accelerate the loan approval process. According to the method, the user data of the user in the loan application process can be collected in real time, and the user data is analyzed to obtain the analysis result, so that the loan limit and interest rate are adjusted, a loan approval decision is made, and the loan approval process is accelerated.

Description

Technical Field

[0001] The present invention relates to the field of information processing technology, and in particular to an information processing system, method, device and medium based on a bank loan platform. Background Art

[0002] With the rapid development of digital financial technology, online loans have become an important trend. You can complete all the steps of loan application without leaving home, including understanding the application conditions of various types of loans, preparing application materials, and submitting loan applications, all of which can be completed efficiently on the Internet. Users first submit their applications online, and consumer finance companies conduct preliminary screening and review. For some applications with higher risks or requiring further verification, offline face-to-face interviews, home visits and other links will be arranged to more accurately assess the risks. In the above-mentioned online application and offline review method, there are many manual intervention links in some approval processes, which are easily affected by subjective factors, such as the experience, emotions, and prejudices of the reviewers, which affect the objectivity and fairness of the approval results. The manual review method is also less efficient. Therefore, there is an urgent need for an information processing solution that can improve the accuracy and efficiency of loan approval. Summary of the invention

[0003] Therefore, the technical problem to be solved by the present invention is to overcome the problems of low accuracy and low efficiency of manual review in related technologies.

[0004] In order to solve the above technical problems, the present invention provides an information processing system based on a bank loan platform, comprising:

[0005] The data embedding design module is used to determine the data embedding plan based on the business process and user behavior characteristics of the front-end H5 page of the bank loan platform; the business process includes the loan application process;

[0006] The data collection module is configured in the front-end H5 page, and is used to collect user data in real time according to the data embedding scheme determined by the data embedding design module, and send the user data to the back-end server;

[0007] The user data includes the user's behavior data and application information; the behavior data includes: the user's browsing path, dwell time and click events; the application information includes: application time, loan amount, loan term, repayment method, personal information: user's name, age and gender, business information: business size and business information, credit history: historical loan records and historical repayment records;

[0008] A data processing and analysis module, configured on the back-end server, for receiving, processing and analyzing the user data to determine a data analysis result;

[0009] The result application module includes a loan approval optimization unit, which is used to optimize the loan approval process according to the data analysis results, including: adjusting the loan amount and interest rate, making loan approval decisions, so as to speed up the loan approval process.

[0010] In an optional implementation, the data embedding design module includes:

[0011] A tracking point demand determination unit is used to determine the tracking point demand according to the business process and user behavior characteristics of the front-end H5 page of the bank loan platform; the tracking point demand includes key nodes and data fields corresponding to the key nodes;

[0012] The tracking point scheme design unit is used to determine the tracking point location according to the key nodes in the tracking point requirements, determine the data type according to the data fields corresponding to the key nodes, and design the corresponding collection frequency.

[0013] In an optional implementation, the data acquisition module includes:

[0014] A data collection unit is used to collect user data in real time according to the data embedding scheme determined by the data embedding design module; a data encryption unit is used to encrypt the user data to obtain encrypted user data; a data compression unit is used to compress the encrypted user data to obtain compressed user data; a data transmission unit is used to send the compressed user data to the back-end server.

[0015] In an optional implementation, the data processing and analysis module includes:

[0016] A data processing unit is used to receive the user data and perform data processing on the user data to obtain the processed user data; the data processing includes: preprocessing and feature extraction; a data analysis unit is used to analyze the processed user data to obtain the data analysis result.

[0017] In an optional implementation, the data analysis unit includes:

[0018] A model building unit, used to build a risk assessment model;

[0019] The model application unit is used to input the processed user data as input data into the risk assessment model to obtain the risk score corresponding to the user; the risk score is used to characterize the credit risk and loan risk of the user.

[0020] In an optional implementation, the loan approval optimization unit is specifically used to adjust the loan amount and interest rate according to the risk score corresponding to the user, and automatically or manually make a loan approval decision based on the risk score corresponding to the user and preset approval criteria.

[0021] In an optional implementation, the data analysis unit further includes:

[0022] The user portrait building unit is used to analyze the user data after data processing and build a user portrait.

[0023] In a second aspect, the present invention provides an information processing method based on a bank loan platform, comprising:

[0024] Determine the data embedding plan based on the business process and user behavior characteristics of the front-end H5 page of the bank loan platform; the business process includes the loan application process;

[0025] Collect user data in real time according to the data embedding solution, and send the user data to the backend server; the user data includes user behavior data and application information; the behavior data includes: user's browsing path, stay time and click events; the application information includes: application time, loan amount, loan term, repayment method, personal information: user's name, age and gender, business information: business size and business information, credit history: historical loan records and historical repayment records;

[0026] Receive, process and analyze the user data to determine data analysis results;

[0027] Optimizing the loan approval process based on the data analysis results includes: adjusting the loan amount and interest rate, and making loan approval decisions to speed up the loan approval process.

[0028] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the information processing method based on the bank loan platform of the second aspect mentioned above by executing the computer instructions.

[0029] In a fourth aspect, the present invention provides a computer-readable storage medium, a single computer-readable storage medium storing computer instructions, and the computer instructions are used to enable a computer to execute an information processing method based on a bank loan platform according to any embodiment of the second aspect.

[0030] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions, wherein the computer instructions are used to enable a computer to execute the information processing method of the second aspect.

[0031] The technical solution provided by the present invention has the following technical effects:

[0032] The technical solution of the embodiment of the present invention can determine the data embedding scheme according to the business process and user behavior characteristics of the H5 page on the front end of the bank loan platform through the data embedding design module. This can accurately collect various data of users in the loan application process, including personal information (personal information, business information, credit history), behavioral data (browsing path, dwell time, click events) and device information (device model, operating system, network environment). Compared with manual review, which may rely on limited information and subjective judgment, this data collection method is more comprehensive and objective, and provides a rich data foundation for accurate evaluation.

[0033] The data collection module can collect user data in real time and send it to the back-end server. This means that the data is highly timely, avoiding the delays and untimely information updates that may occur during manual data collection, so that the back-end data processing and analysis modules can perform analysis based on the latest data, improving the accuracy of the analysis results.

[0034] The data processing and analysis module receives and processes user data on the back-end server and quickly obtains data analysis results. This process avoids the tedious process of manually checking and analyzing user data one by one, greatly improving efficiency.

[0035] The loan approval optimization unit in the result application module optimizes the loan approval process based on the data analysis results, such as adjusting the loan amount and interest rate and making loan approval decisions. This data-driven decision-making method can speed up the loan approval process, reduce the repeated communication and information supplementation that may occur during the manual review process, and improve the overall approval efficiency.

[0036] Through comprehensive and real-time data collection and automated data analysis, more accurate information can be provided for loan approval, reducing inaccuracies caused by factors such as limited information and subjective judgment during manual review, thereby effectively solving the problem of low accuracy of manual loan review. Automated data processing and data-based approval process optimization reduce manual intervention links, speed up the approval process, and solve the problem of low efficiency of manual review. Therefore, the technical solution of the present invention can solve the technical problems of low accuracy and low efficiency of manual loan review to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the related technologies, the drawings required for use in the specific embodiments or the related technical descriptions will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1It is a structural schematic diagram of an information processing system based on a bank loan platform according to an embodiment of the present invention;

[0039] Figure 2 It is a schematic diagram of a data embedding solution according to an embodiment of the present invention;

[0040] Figure 3 is a schematic diagram of a visual analysis result of an embodiment of the present invention;

[0041] Figure 4 It is a flowchart of an information processing method based on a bank loan platform according to an embodiment of the present invention;

[0042] Figure 5 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0044] To this end, the embodiments of the present invention provide an information processing system, method, device and medium based on a bank loan platform to solve the problems of low accuracy and low efficiency of manual review.

[0045] According to an embodiment of the present invention, an information processing system embodiment based on a bank loan platform is provided. It should be noted that a single system is used to implement the following embodiments and optional implementation methods. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the system described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0046] Figure 1 FIG. 1 is a schematic diagram of another information processing system based on a bank loan platform according to an embodiment of the present invention. Figure 1 As shown, an embodiment of the present invention provides an information processing system based on a bank loan platform, and the information processing system based on a bank loan platform includes:

[0047] Data embedding design module 11, data collection module 12, data processing and analysis module 13 and result application module 14.

[0048] The data embedding design module 11 includes: a embedding requirement determination unit 111 and a embedding solution design unit 112 .

[0049] The data collection module 12 includes: a data collection unit 121 , a data encryption unit 122 , a data compression unit 123 , and a data transmission unit 124 .

[0050] The data processing and analysis module 13 includes: a data processing unit 131 and a data analysis unit 132. The data analysis unit 132 includes: a model building unit 1321 and a model application unit 1322. The data analysis unit 132 also includes: a user portrait building unit 1323.

[0051] The result application module 14 includes a loan approval optimization unit 141 .

[0052] The data embedding design module 11 is used to determine the data embedding scheme according to the business process and user behavior characteristics of the front-end H5 page of the bank loan platform. The business process includes the loan application process. The business process may also include: product browsing and comparison process, user authentication and authorization process, consultation and customer service process, repayment management process, etc. As an example, the data embedding scheme is as follows Figure 2 shown.

[0053] Product browsing and comparison process:

[0054] Product display page browsing: When users enter the loan product display page, record the types of loan products browsed by users (such as personal consumption loans, housing loans, business loans, etc.) and the browsing time (stay time) of each product. This can help understand the user's initial interest in different loan products. For example, if users stay on the housing loan product page for a long time, it may mean that they have potential demand for home purchase loans.

[0055] Product details viewing: Track the behavior of users clicking to view loan product details and record the details of the products viewed, including interest rate, loan term, repayment method, loan amount range, etc. These data can be used to analyze the characteristics of loan products that users focus on, and provide a basis for precision marketing and product optimization. For example, if many users frequently check whether there is a penalty clause for early repayment of a certain product, this point may need to be explained more clearly in the product introduction.

[0056] Product comparison behavior: When users use the product comparison function provided by the platform (if any), record the product combination being compared and the dimensions of comparison (such as interest rate, amount flexibility, etc.). This helps to understand the trade-offs that users make when choosing loan products so as to adjust product strategies.

[0057] 2. User authentication and authorization process:

[0058] Authentication steps: When users perform identity authentication, such as uploading a photo of their ID card, entering their mobile phone number to obtain a verification code, and face recognition (if any), record the completion of each step, whether errors or repeated operations occurred. This data is very important for optimizing the authentication process, improving user experience, and preventing identity authentication risks. For example, if many users fail multiple times in the face recognition step, it may be necessary to check whether there are technical problems with the function or provide clearer operating instructions.

[0059] Authorization operation: Track the user's authorization to the platform to query their credit information, bank statements, etc., including whether the authorization is successful or not, the time of authorization, etc. Understanding the user's acceptance of the authorization operation and possible problems will help improve authorization reminders and instructions, and also better track the process of obtaining user credit data.

[0060] 3. Consultation and customer service process:

[0061] Online consultation initiation: Record the entry point where the user initiates the online consultation (such as on the product details page, during the application process, or a dedicated customer service page), and the type of question asked (such as product details, application conditions, interest rate calculation, etc.). This can help understand the user's confusion points in the loan process in order to optimize product introductions and customer service knowledge bases. Customer service communication details: If the platform has a chat-based customer service function, record the chat content, chat duration, problem solving status, etc. between the customer service and the user. These data help evaluate the quality of customer service, and can also discover common user questions for optimizing the platform's FAQ section.

[0062] Customer service evaluation feedback: collect user evaluations of customer service, including evaluation levels (such as satisfied, average, dissatisfied) and specific evaluation content. This is crucial to improving customer service quality and increasing user satisfaction.

[0063] 4. Repayment management process (for users who have taken out loans):

[0064] Repayment plan review: When a user who has taken out a loan logs into the platform to review the repayment plan, the frequency of review, the specific time of review (such as when the payment is close to the due date, a long time in advance, etc.), and whether the repayment plan has been printed or downloaded are recorded. This can help understand the user's attention to the repayment plan and management habits.

[0065] Repayment reminder interaction: For repayment reminder messages (SMS, push notifications, etc.) sent by the platform, record whether the user has viewed the reminder and the actions taken after viewing it (such as ignoring, clicking to enter the repayment page, etc.). These data help optimize the repayment reminder strategy and improve the repayment rate.

[0066] Repayment operation records: When users make repayments (such as online payment repayments, setting automatic repayments, etc.), the repayment method, whether the repayment is successful, whether the repayment is made in advance, and other details are recorded. This is of great significance for evaluating the user's repayment ability and credit status.

[0067] In this embodiment, the data embedding design module 11 may include:

[0068] The embedding demand determination unit 111 is used to determine the embedding demand according to the business process and user behavior characteristics of the front-end H5 page of the bank loan platform. The embedding demand includes key nodes and data fields corresponding to the key nodes.

[0069] The tracking solution design unit 112 is used to determine the tracking location according to the key nodes in the tracking requirements, determine the data type according to the data fields corresponding to the key nodes, and design the corresponding collection frequency.

[0070] Difficulties in building traffic data in traditional technologies:

[0071] The quality of buried points is poor: wrong buried points or missed buried points, the quality is uneven, and the use of data is risky.

[0072] The non-standard design of tracking points leads to discrete and chaotic tracking points, which is not conducive to unified use. There is a lack of consent management for various types of data, making it difficult to realize the value of data.

[0073] If business utilization is low and not used, it will consume costs.

[0074] The pain points of various roles in the previous bank loan platform in traffic data practice:

[0075] Data users (business departments): There is no unified management of data points. The logic is meaningless, and the attribute content and responsible person are missing.

[0076] Tracking point developers: There are no records of the addition, modification, abandonment, and decommissioning of tracking points. We dare not touch the past tracking point codes, and the code is becoming more and more confusing.

[0077] The party that demands the tracking points (the operator of the bank loan platform, the pre-loan, mid-loan and post-loan teams): The tracking point demands have nowhere to be maintained, and flexible and controlled version management is impossible.

[0078] Data manager (operator of the bank loan platform): What is the current status of traffic data assets? There is a lack of governance and management tools for the number of embedding points, data volume, cost consumption, number of effective embedding points, etc.

[0079] The tracking demand determination unit is specifically used to decompose the loan application process: the loan application process is subdivided into multiple sub-processes, such as user registration and login, filling in personal information (personal information, corporate information, etc.), uploading information (ID card photo, business license, etc.), selecting loan product type and amount, submitting application for review, etc.

[0080] Analyze the user's interactive behavior in each sub-process, such as the user's input behavior and operations on required and optional items when filling in personal information. When uploading materials, file selection, feedback processing on upload success or failure, etc.

[0081] Identification of key behaviors: Identify user behaviors that have a significant impact on loan approval, such as the length of time spent browsing the loan product details page (reflecting the user's attention to and understanding of the product), the number of modifications made when filling in credit history information (which may indicate the authenticity and stability of the information), the number of times the entire application page is repeatedly viewed before submitting the application (reflecting the user's level of caution), etc.

[0082] Determine the tracking needs: Based on the above business process decomposition and key behavior analysis, determine the key nodes of tracking. For example: Registration login page: login button click, registration success event, etc. Personal information filling page: name, age, gender and other fields start input event, complete input event, save information event. Business information page: business size field input, business information text box content change event, etc. Credit history page: historical loan record query button click, historical repayment record display event, etc. Loan product selection page: product click event, credit limit selection event, etc. Application submission page: submit button click event, application submission success event, etc.

[0083] Determine the corresponding data fields: For each key node, determine the data fields that need to be collected. For example: Registration and login page: user name, password, registration time, login time, login IP address, etc. Personal information filling page: name, age, gender, ID number, contact information, etc. Business information page: enterprise scale (number of employees, registered capital and other specific data), business scope (text description), enterprise establishment time, etc. Credit history page: historical loan amount, loan term, repayment method, repayment time, whether there is an overdue record, etc. Loan product selection page: product name, product number, selected amount, interest rate type, etc. Application submission page: application time, application number, application status, etc.

[0084] The tracking plan design unit is used to determine the tracking location, data type and collection frequency.

[0085] Determine the location of the tracking point: For the front-end H5 page, the tracking point location is usually in the JavaScript function triggered by the relevant event. For example, to track the click event of the login button, insert the data collection code in the click response function of the login button. For data collection when the page is loaded (such as obtaining device information when the page is loaded for the first time), track the data in the `onload` event function of the page.

[0086] Data type determination: Determine the data type based on the nature of the data field. For example, text information such as name and company name is of string type. Age, number of employees, loan amount, etc. are of numeric type. Registration time, login time, repayment time, etc. are of date and time type. Device model, operating system, etc. are of string type (the enumeration values ​​of device type and operating system can be pre-defined to facilitate subsequent analysis). Whether there is an overdue record, etc. is of Boolean type (true if there is an overdue record, false if there is no overdue record).

[0087] Determination of collection frequency: For some one-time events, such as successful registration, successful application submission, etc., the collection frequency is 1 time, that is, the collection is performed when the event occurs. For the user's continuous behavior data on the page, such as browsing paths, a higher collection frequency can be set, such as collecting the URL information of the user's current page every 1 second to fully record the user's browsing trajectory. For changes in the content of the user information input field, it can be collected every time the content changes to ensure the real-time and integrity of the data. As for device information, since it is basically determined when the user enters the page and does not change frequently, it can be collected once when the page loads.

[0088] Point embedding location: Home page: Button clicks: such as the "Apply Now" button and the "Learn More" button, etc., to record the user's entry intention. Page dwell time: record the user's dwell time on the home page to evaluate the attractiveness of the home page. Application process page: Form filling: input events for each input box, including input content, input time, etc., to understand the fluency of user filling and possible problems encountered. Drop-down menu selection: record the user's selection when selecting drop-down menus such as loan term and repayment method. Submit button: record the time and number of times the user submits the application. Result page: Approval result display: record the approval results (passed, rejected, etc.) seen by the user and the time to view the results. Buttons that guide the next step: such as clicks on buttons such as "Reapply" and "View Details".

[0089] Data types: 1) Behavioral data: A. Click events: record the buttons, links and other elements clicked by users. B. Page browsing events: record the time when users enter and leave each page. C. Form filling data: including input content, selected options, etc. 2) Status data: A. Network status: record the network status of users during operation, such as good, poor, etc. 3) Device information: including device type, operating system version, etc. 4) Business data: information such as loan application amount, term, purpose, etc., and 5) Approval result status (approved, failed).

[0090] Collection frequency: 1) Real-time collection: Important behaviors such as button click events and page switching events are collected in real time to understand user operations in a timely manner. 2) Scheduled collection: The page dwell time can be collected every certain time (such as 10 seconds), and the total dwell time is finally calculated. During the form filling process, the input status can be collected every certain time (such as 30 seconds) to prevent data loss.

[0091] The data collection module 12 is configured in the front-end H5 page, and is used to collect user data in real time according to the data embedding plan determined by the data embedding design module, and send the user data to the back-end server.

[0092] In this embodiment, after the user data is sent to the backend server, the user data collected in real time may be screened, and user data satisfying the filtering conditions may be screened by inputting the filtering conditions.

[0093] In this embodiment, user data includes but is not limited to user behavior data and application information. Behavior data includes: user's browsing path, dwell time and click events. Application information includes: application time, loan amount, loan term, repayment method, personal information: user's name, age and gender, business information: financial status such as business size and business information, credit history: historical loan records and historical repayment records.

[0094] Design an indicator system for data collection, including user personal information, behavior data, device information, etc. Personal information can include the user's name, age, gender, etc. Behavior data can include the user's browsing path, dwell time, click events, etc. Device information can include the user's device model, operating system, network environment, etc. By comprehensively collecting these indicators, ensure the comprehensiveness and accuracy of the data.

[0095] In this embodiment, the user data also includes device information, and the device information includes: the user's device model, operating system, and network environment.

[0096] In this embodiment, the data acquisition module 12 may include:

[0097] The data collection unit 121 is used to collect user data in real time according to the data embedding plan determined by the data embedding design module.

[0098] In this embodiment, the data collection unit 121 is specifically used to collect data at the buried point location according to the corresponding collection frequency and data type.

[0099] The data encryption unit 122 is used to encrypt the user data to obtain encrypted user data.

[0100] In this embodiment, the user data can be encrypted by conventional technical means in the art to obtain encrypted user data. For example, symmetric encryption algorithm, asymmetric encryption algorithm, hash encryption, etc. As an example, the AES algorithm is used for encryption: determine the key length, AES supports 128-bit, 192-bit and 256-bit keys, the longer the key length, the higher the security, but the encryption and decryption speed may be slightly slower. For example, a 128-bit key can be selected, and a key with a length of 16 bytes (128 bits converted to bytes) can be generated by a secure random number generator. Encryption process: After the user data is obtained on the front-end H5 page or the back-end server (depending on the specific encryption implementation location), the data is encrypted using the selected AES algorithm and key. For example, in the JavaScript environment (front-end), some mature encryption libraries (such as CryptoJS) can be used to implement it. After the back-end server receives the encrypted user data (assuming that it is transmitted in Base64 encoding), it is necessary to first perform Base64 decoding and then use the same key to decrypt.

[0101] The data compression unit 123 is used to compress the encrypted user data to obtain compressed user data.

[0102] In this embodiment, the encrypted user data may be compressed using conventional technical means in the art to obtain compressed user data, for example, a lossless compression algorithm, such as the GZIP algorithm.

[0103] The data transmission unit 124 is used to send the compressed user data to the backend server.

[0104] In this embodiment, the compressed user data may be sent to the backend server using conventional technical means in the art, for example, using HTTP protocol, message queue, etc.

[0105] In this embodiment, in the front-end H5 page of the bank loan platform, the real-time collection of user behavior is realized by introducing the data embedding SDK. The data embedding SDK can collect the user's click, input, sliding and other behavior data on the page, and send it to the back-end server for processing in real time.

[0106] The data processing and analysis module 13 is configured in the back-end server and is used to receive, process and analyze user data to determine data analysis results.

[0107] In this embodiment, the data processing and analysis module 13 may include:

[0108] The data processing unit 131 is used to receive user data and process the user data to obtain the processed user data. Data processing includes: preprocessing and feature extraction. Preprocessing includes data cleaning: removing invalid, erroneous or duplicated data, formatting and normalization. The collected user data is cleaned to remove invalid, erroneous or duplicated data. For example, if the user age field has an obviously unreasonable value (such as over 150 years old), the record is corrected or deleted. The user data is formatted and normalized to make it suitable for subsequent analysis. Feature extraction: extract features from user data, such as the user's browsing path, operation frequency, standardization of input content, etc., and combine the user's application information, such as basic information, financial data, etc., to build a more comprehensive feature set. Select features that have a significant impact on risk assessment from a large number of user data, such as age and income level in personal information, number of overdue payments and historical loan times in credit history, and frequency of recent browsing of loan products in behavioral data. At the same time, some features can also be combined or new features can be derived, such as combining age and income level to generate a comprehensive indicator reflecting repayment ability. For categorical features (such as gender, occupation, etc.), they need to be encoded and converted into numerical form for model processing. Common encoding methods include one-hot encoding, etc. For numerical features, in order to make different features in the same dimension and facilitate model learning, normalization processing can be performed, such as mapping the value to the 0-1 interval or performing standardization processing (making its mean 0 and variance 1).

[0109] The data analysis unit 132 is used to analyze the processed user data to obtain data analysis results.

[0110] In this embodiment, the data analysis unit 132 may include: a model building unit 1321 and a model application unit 1322 .

[0111] The model building unit 1321 is used to build a data analysis model, which includes a risk assessment model, a fraud detection model, an approval efficiency prediction model, etc.

[0112] In this embodiment, data analysis models can be constructed using machine learning, data mining and other technologies. Data analysis models include but are not limited to risk assessment models, fraud detection models, approval efficiency prediction models and the like.

[0113] Real-time data analysis through data analysis models: Quickly analyze the real-time incoming data and calculate key indicators such as user risk score, fraud risk, etc. Analyze bottlenecks and potential problems in the approval process, such as some links taking too long and high error rates in user information.

[0114] Dynamically adjust policies through the result application module: Dynamically adjust approval processes and policies based on data analysis results. For example, for users with high risk scores, the approval process can be simplified and the approval time can be shortened. For applications with high fraud risks, the review can be strengthened or the application can be rejected.

[0115] For the risk assessment model, in this embodiment, the user data after data processing is divided into a training set, a validation set and a test set according to a certain ratio. Usually, the training set is used to train the model so that the model learns the relationship between features and risks. The validation set is used to adjust the hyperparameters of the model during the model training process and select the optimal model configuration. The test set is used to finally evaluate the performance of the model on unseen data. The common division ratio can be 7:2:1 (training set: validation set: test set). Training with training set: Taking the logistic regression model as an example, the divided training set data (assuming that the feature data is stored in the X_train matrix, the corresponding risk label data is stored in the y_train vector, and the risk label can be 0 for low risk and 1 for high risk) is used for training. Hyperparameter adjustment (optional): For some models (such as the maximum depth of the decision tree, the number of trees in the random forest, and other parameters), it is necessary to adjust the hyperparameters through the validation set to find the optimal parameter combination to make the model perform best. Common hyperparameter adjustment methods include grid search (GridSearch), random search (RandomSearch), etc. Model evaluation: Use the test set data (feature data is X_test, and the corresponding risk label is y_test) to evaluate the trained model. Common evaluation indicators include accuracy, precision, recall, F1 value, and area under the ROC curve (AUC). Risk score acquisition and application: The user data that needs to be evaluated in the actual application (after the same feature engineering processing as before) is used as input into the trained risk assessment model. The model will output the corresponding prediction results (for classification models, such as 0 or 1 to represent the risk category. For models that can output probabilities, such as the probability value between 0 and 1 output by logistic regression, which can be used as a risk score, the higher the probability, the greater the risk). The credit risk and loan risk of the user are characterized and decided based on this risk score. For example, a threshold is set (such as 0.5). Users with risk scores greater than the threshold are considered high-risk users, and more cautious measures may be taken in loan approval (such as reducing the amount, increasing the interest rate, or directly rejecting the loan application, etc.).

[0116] The model application unit 1322 is used to input the processed user data as input data into the risk assessment model to obtain the risk score corresponding to the user. The risk score is used to characterize the credit risk and loan risk of the user.

[0117] In this embodiment, the risk score calculation includes:

[0118] User data collection: Collect multi-dimensional data of users, including but not limited to personal information (such as age, occupation, income, etc.), business information (such as business size, revenue, profitability, etc.), credit history (such as past loan records, repayment status, etc.), and behavioral data (such as operating behavior on the H5 page, dwell time, etc.).

[0119] Indicator setting: Determine the indicator weights corresponding to each data. For example, credit history may have a higher weight because it can directly reflect the user's repayment ability and credit status. Business information will also have a certain weight to reflect the user's economic strength and stability.

[0120] Scoring model: Use mathematical models (risk assessment models) to comprehensively calculate various indicators. Common models include linear regression, logistic regression, decision trees, etc., which can be used as risk assessment models. Through these models, the values ​​of different indicators are converted into a specific risk score.

[0121] In this embodiment, risk assessment model construction: a machine learning algorithm can be used to train a risk assessment model based on user data. The risk assessment model will learn how to predict the user's loan risk based on the features.

[0122] Real-time analysis: When users operate on the H5 page, the server processes the user data and behavior data in real time, and uses the trained model to perform risk assessment. The model outputs a risk score that reflects the user's credit risk and loan risk.

[0123] In this embodiment, the data analysis unit 132 may further include: a user portrait construction unit 1323, which is used to analyze the processed user data and construct a user portrait.

[0124] The following are the detailed steps to build a user portrait:

[0125] 1. Data collection and integration:

[0126] 1. Personal information collection: Age: obtained from user registration information or real-name authentication channels. People of different ages often have different consumption and loan needs. For example, young people may prefer consumer loans for education or electronic products, while middle-aged people may consider housing loans or commercial investment loans.

[0127] Occupation: The occupation information is supplemented and improved through the work unit and position information filled in by the user, or with the help of a third-party data platform. Occupation is closely related to income stability, loan purpose and repayment ability. For example, civil servants and employees of large enterprises usually have a relatively stable source of income, which may have certain advantages in loan approval. However, the income of freelancers or people in the early stages of entrepreneurship fluctuates greatly, and risk assessment requires comprehensive consideration of multiple factors.

[0128] Income: This can be reported by the user, but needs to be verified in combination with other data. Some bank loan platforms may require users to provide information such as salary flow and tax payment certificates to more accurately determine their income level, thereby assessing their loan repayment ability and credit risk. For example, high-income users may have higher loan amount requirements and stronger repayment ability, but may also face more investment temptations and potential risks.

[0129] 2. Business information collection (for business owners or related users) Enterprise size: determined based on the number of employees, asset size, office space area and other information of the enterprise. Large enterprises may have advantages in financing channels and costs, and their loan needs are often related to strategic decisions such as expansion and mergers and acquisitions. Small and micro enterprises may pay more attention to working capital loans to maintain daily operations, and due to their small size, their risk resistance is relatively weak. When assessing risks, it is necessary to focus on their operating stability and cash flow. Revenue and profitability: data is obtained through corporate financial statements, tax return records or docking with corporate financial software. Enterprises with stable revenue growth and good profitability are usually regarded as low-risk customers when applying for loans and may obtain more favorable loan terms. On the contrary, enterprises with declining revenue or long-term losses may face stricter scrutiny for their loan applications, and banks may require additional guarantees or collateral.

[0130] 3. Credit history Collection of past loan records: Query all historical loan information of users from the credit reporting system, including lending institutions, loan amounts, loan terms, repayment status, etc. Users who have repaid their loans on time and in full many times have higher risk scores and lower credit risks. Users with overdue records, especially those with more overdue times or longer overdue periods, have significantly increased credit risks, and banks may increase interest rates, reduce loan amounts, or directly reject applications when approving loans. Repayment status: Analyze the repayment details of each loan in detail, such as whether there are early repayments, partial repayments, and the frequency and duration of overdue repayments. Early repayments may reflect that the user has sufficient funds or is sensitive to interest rates. Frequent overdue repayments indicate that there are problems with the user's repayment ability or willingness, which is an important negative indicator for credit risk assessment.

[0131] 4. Behavioral data collection (based on H5 pages) Operational behavior: record the user's clicks, slides, submissions and other operations on the H5 page. For example, if a user frequently clicks on a high-amount loan product page, it may imply that he has a large demand for funds. In-depth browsing of the loan product details page, including checking the interest rate, repayment method, term and other information, shows that the user has a high degree of attention and willingness to understand the loan product. In subsequent marketing or risk assessment, relevant information can be provided or the review can be strengthened. Stay time: Count the user's stay time on different pages or specific functional modules. Staying on the credit assessment prompt page for a long time may mean that the user is more concerned about his own credit status or is carefully checking the information. Staying on the loan application submission page for too long without submitting may be hesitant about the loan decision. Further analysis of the reasons is needed, such as whether there are difficulties in filling in the information, doubts about the loan terms, etc., so as to provide corresponding help or optimize the page design.

[0132] 2. Data cleaning and preprocessing:

[0133] 1. Missing value processing For missing data in personal information, business information or credit history, first determine the cause of the missing and the importance of the data. If key data is missing (such as income information is missing and cannot be verified by other means), it may be necessary to further communicate with the user to obtain supplementary information, or make reasonable guesses or fill in based on other relevant information of the user. For example, for users who are teachers and have a long working life, the average income level of teachers with the same title in the same region can be used for filling. For non-critical data missing (such as missing detailed house numbers in the company's registered address), the majority can be used for filling (if there are many identical values ​​in the field) or the missing value can be directly ignored. The specific processing method depends on the usage scenario of the data and its impact on the accuracy of the model.

[0134] 2. Processing of outliers Check out outliers in personal information, such as age beyond a reasonable range (greater than 120 years old or less than 0 years old), income that is too high or too low (seriously inconsistent with the occupational or regional average), etc. For outliers, further verification of their authenticity is required. If the user fills in the wrong information or the data is entered incorrectly, it should be corrected in time. If it is a special case (such as a high-income celebrity or corporate executive), relevant supporting materials need to be collected for confirmation. In the business information of an enterprise, if there is an abnormal and large fluctuation in revenue or profit, it is necessary to deeply analyze the reasons behind it. It may be that the enterprise has made major strategic adjustments, sudden changes in the market, or financial data fraud, etc. The reliability of the data is ensured by comparing and verifying with multi-channel data such as corporate financial statements and industry reports. For abnormal overdue records in credit history (such as a loan that has been overdue for several years and has not been processed, but the user's other loan records are good), it is necessary to investigate whether there are special disputes or system errors, and make corresponding corrections or annotations.

[0135] 3. Data standardization and normalization For numerical data, such as age, income, corporate revenue, etc., standardization is performed to make it have a unified dimension and distribution characteristics. Common standardization methods include Z-score standardization, which converts the data into a standard normal distribution with a mean of 0 and a standard deviation of 1 by calculating the difference between each data point and the mean and dividing it by the standard deviation. This can avoid the adverse effects on model training due to large differences in numerical values ​​between different features. For some data with a clear range of values ​​or proportional relationships, such as risk scores (usually between 0-100) or the frequency of page operations (values ​​between 0-1 represent relative proportions), normalization can be used to map the data to a specific interval to facilitate the model to understand and compare the importance of different features.

[0136] 3. User portrait label construction:

[0137] 1. Basic attribute labels Age label: divided according to age range, such as 20-30 years old for young people, 31-50 years old for middle-aged, and 51 years old and above for the elderly. Different age labels can be further subdivided, such as young people can be divided into 20-25 years old (just entering the society), 26-30 years old (career start-up period), etc., so as to more accurately analyze the loan needs and risk characteristics of users of different age groups. Occupation label: classified according to industry category (such as finance, Internet, manufacturing, etc.) and occupation type (such as managers, technicians, ordinary employees, etc.). For example, financial industry practitioners may have a deeper understanding of financial products and may pay more attention to interest rate concessions and flexible repayment methods in loan selection. Ordinary employees in the manufacturing industry may pay more attention to the amount and approval speed of the loan to meet the needs of living consumption or house and car purchase. Income level label: users are divided into low-income, middle-income, and high-income groups according to local income levels and industry average income. For example, in first-tier cities, an annual income of less than 100,000 yuan may be considered low income, 100,000-500,000 yuan is middle income, and more than 500,000 yuan is high income. The income level label is directly related to the loan amount and repayment ability, and is an important basis for credit risk assessment and loan product recommendations.

[0138] 2. Credit risk label Risk score label: Based on the user's credit history information, the user's risk score is calculated through the risk score model, and the credit level is divided accordingly, such as excellent (80-100 points), good (60-79 points), medium (40-59 points), and poor (0-39 points). The risk score label can intuitively reflect the user's overall credit status. The higher the credit level, the higher the loan approval rate and the lower the interest rate. Conversely, users with low credit levels may face loan restrictions or high-cost loans. Overdue risk label: Based on factors such as the number of overdue loans, overdue duration, and overdue amount of the user's past loans, the user's overdue risk level is assessed and divided into low risk (no overdue or occasional short-term overdue), medium risk (a certain number of overdue records but has been paid off or is improving), and high risk (multiple overdue with large overdue amounts and long overdue periods). Overdue risk labels can help banks formulate risk prevention measures in advance, such as requiring high-risk users to provide additional guarantees or increase the down payment ratio.

[0139] 3. Behavioral characteristic labels Activity labels: Based on the user's login frequency, number of operations and dwell time on the H5 page, users are divided into high-active, medium-active and low-active users. Highly active users may have a high degree of attention and demand for loan products, and can be used as key marketing targets to regularly push personalized loan product information and promotions. Low-active users need to further analyze the reasons for their inactivity, such as whether the platform experience is poor, the product does not meet the needs, etc., so as to take corresponding improvement measures. Product preference labels: Determine the user's product preference by analyzing the user's click and browsing behavior on the page for different loan product types (such as consumer loans, housing loans, commercial loans, etc.) and product features (such as interest rates, repayment periods, amount, etc.). For example, users who frequently browse housing loan products and pay attention to the product features of low interest rates and long periods can be marked as housing loan preference users, and suitable housing loan products and related services can be recommended in a targeted manner in subsequent marketing.

[0140] 4. Business status label (for business owners) Business size label: According to indicators such as the number of employees and asset size, enterprises are divided into small and micro enterprises, medium-sized enterprises, and large enterprises. Enterprises of different sizes have obvious differences in loan demand, risk tolerance and financing channels. Small and micro enterprises usually rely more on external financing such as bank loans, and the loan amount is relatively small. Large enterprises may have more diverse financing options, and loan demand is more related to strategic expansion and capital operation. Business stability label: According to factors such as the company's revenue growth rate, profit stability, market share changes, etc., the business stability of the company is evaluated and divided into stable growth type, fluctuation type, and decline type. Enterprises with stable growth have more advantages in loan approval and can obtain more favorable loan conditions. Enterprises with declining operations may face higher risk assessments and stricter loan approval processes. Banks may require enterprises to provide detailed business improvement plans and guarantee measures.

[0141] 4. User portrait update and maintenance:

[0142] 1. Regular update mechanism sets a fixed time period (such as monthly or quarterly) to update the user portrait. Over time, the user's personal information, business conditions, credit history and behavioral data may change. For example, a user's income may increase or decrease due to a promotion, salary increase or industry change. The company's revenue and profitability will change with the adjustment of the market environment and business strategy. Credit history will be continuously updated with new loan applications and repayment records. User behavior on the H5 page will also vary due to platform function optimization, product updates or changes in personal needs. By regularly updating the user portrait, the user's latest status can be reflected in a timely manner, ensuring that the bank uses accurate and effective data in business links such as loan approval, risk assessment and product recommendation.

[0143] 2. In addition to regular updates, a real-time update mechanism should be established to update the user portrait immediately when certain key events occur. For example, after a user successfully applies for and obtains a new loan, his credit history and debt situation change, and the relevant portrait tags should be updated in a timely manner. When a company undergoes major equity changes, asset restructuring, or legal proceedings, this information may have a significant impact on the company's operating conditions and credit risk, and the business owner's user portrait also needs to be updated immediately. In addition, some important user behaviors on the H5 page, such as submitting loan applications, modifying personal information, or conducting large-scale fund transactions, can also be used as real-time update triggers so that banks can keep abreast of user dynamics and make corresponding business decisions.

[0144] 3. Data quality monitoring and maintenance During the user portrait update process, strengthen the monitoring and maintenance of data quality. Regularly check the integrity, accuracy and consistency of the data, and promptly identify and handle issues such as missing data, outliers and erroneous data. At the same time, establish a data backup and recovery mechanism to prevent data loss or damage from adversely affecting the construction and application of user portraits. In addition, as the business develops and data sources change, continuously optimize the data collection and processing process, improve data quality and availability, and provide solid guarantees for the accurate construction and effective application of user portraits.

[0145] In this embodiment, the data analysis unit 132 may further include:

[0146] User behavior analysis unit, used to analyze page browsing behavior: track the user's browsing path on the H5 page to understand user preferences and behavior patterns. Analyze operation frequency and dwell time: analyze the frequency and dwell time of users in different operations (such as filling out forms, clicking buttons).

[0147] Device analysis unit for analyzing device type and operating system: Identify the user's device type and operating system to facilitate personalized services. Analyze geographic location: Understand the user's location information through GPS data for risk assessment and market positioning.

[0148] Personal information analysis unit, used to analyze basic user information: analyze personal information submitted by users, such as age, occupation, etc., to assess loan demand. Analyze financial status: analyze financial information submitted by users, such as income, debt, etc., to assess repayment ability.

[0149] The default prediction unit is used to predict the user's future default probability to help banks formulate loan strategies.

[0150] The collected data is processed and analyzed in real time through the server to build a user profile and risk assessment model. The server can analyze the user's loan needs, credit status and other information based on the user's behavior data to provide a basis for loan approval. For example, the user's interests and needs can be analyzed based on the user's browsing path and stay time. The user's operating habits and credit willingness can be analyzed based on the user's click events.

[0151] The result application module 14 includes a loan approval optimization unit 141, which is used to optimize the loan approval process according to the data analysis results, including: adjusting the loan amount and interest rate, making loan approval decisions, so as to speed up the loan approval process.

[0152] In this embodiment, the loan approval optimization unit 141 is specifically used to adjust the loan amount and interest rate according to the risk score corresponding to the user, and automatically or manually make a loan approval decision according to the risk score corresponding to the user and the preset approval criteria.

[0153] In this embodiment, loan approval decision: Loan approval decisions are made automatically or manually based on risk scores and preset approval criteria. High-risk loan applications may be rejected or require additional review and guarantees. Approval tools (such as RPA robotic process automation) can be automated to reduce manual intervention. Strategy optimization: Adjust loan amount, interest rate, term and other strategies based on market changes and user needs.

[0154] The loan approval optimization unit 141 is also used to identify and eliminate redundant links and invalid operations in the approval process, optimize the process sequence and node settings, and improve the approval efficiency.

[0155] In this embodiment, the result application module also includes a monitoring and feedback unit, which is used to monitor the approved loans in real time and detect potential risks in a timely manner. The monitoring results and loan performance feedback are built into the model for continuous optimization of the risk assessment model. The monitoring and feedback unit is also used to establish a risk monitoring mechanism to monitor abnormal situations in the loan approval process in real time. Set the warning threshold, and once the warning condition is triggered, immediately start the emergency response mechanism. Feedback and iteration: User feedback collection: Collect user feedback through user surveys, satisfaction surveys, etc. Analyze user feedback data to understand user needs and pain points. Effect evaluation: Evaluate the effect of the optimized approval process and strategy. Evaluation indicators include but are not limited to approval pass rate, approval time, user satisfaction, etc. Continuous optimization: Continuously optimize the approval process and strategy based on the evaluation results and user feedback.

[0156] In this way, the technical solution of the present invention can achieve real-time assessment of loan risks, improve approval efficiency and accuracy, and reduce default risks.

[0157] In an optional implementation, the result application module 14 also includes: a product optimization unit for optimizing the front-end H5 page of the bank loan platform in terms of product function, design, marketing strategy, etc. based on the data analysis results.

[0158] The user experience improvement unit is used to optimize the user experience and improve user satisfaction and conversion rate based on user behavior data.

[0159] The risk control unit is used to identify and warn of potential risks, reduce bad debt rates, and ensure the sound operation of banking business.

[0160] In an optional implementation, the data embedding design module 11 further includes:

[0161] The custom event tracking unit is used to customize the tracking of specific events according to business needs to capture user behavior data in specific scenarios.

[0162] The tracking verification unit is used to verify the correctness and validity of the tracking code to ensure the accuracy and completeness of data collection.

[0163] In an optional implementation, the data processing and analysis module 13 further includes:

[0164] The data visualization unit is used to visualize the analyzed data in the form of charts, reports, etc., to facilitate business personnel to understand and apply.

[0165] The trend prediction unit predicts future user behavior and market trends based on historical data and behavior patterns, providing data support for business decisions.

[0166] In an optional implementation, the information processing system based on the bank loan platform further includes:

[0167] The user privacy protection module is used to desensitize user sensitive information during data collection, processing and storage to ensure the security and compliance of user privacy.

[0168] The permission management module is used to set the access and operation permissions of different users to system functions to ensure data security and system stability.

[0169] In an optional implementation, the data embedding design module 11 and the data acquisition module 12 are implemented through a front-end framework or SDK (software development kit) to facilitate system integration and maintenance.

[0170] The data processing and analysis module 13 uses big data processing technology and machine learning algorithms to improve the efficiency and accuracy of data analysis.

[0171] In an optional implementation, the system can monitor the operation of data points in real time, including data collection, transmission, processing and storage, and promptly detect and handle abnormal situations to ensure system stability and data reliability.

[0172] In an optional implementation, the information processing system based on the bank loan platform also includes: an alarm notification module, which is used to send alarm notifications to relevant personnel by email, text message, etc. when abnormal situations are found, so as to facilitate timely processing. A log recording module is used to record the system operation log and key information in the data processing process to facilitate the tracing and troubleshooting of problems.

[0173] The embodiment of the present invention also provides a full-process data embedding service based on the front-end H5 page of the bank loan platform, which provides a full-process service from data embedding design, data collection, processing and analysis to result application, including but not limited to:

[0174] Provide consulting and solution development services for data embedding design. Provide front-end SDK or framework integration and maintenance services. Provide back-end data processing and analysis services, including data cleaning, integration, visualization and trend forecasting. Provide advice and strategic support services for business optimization and risk control.

[0175] The technical solution of the present invention realizes the real-time collection, transmission and analysis of the user's behavioral data in the application process of the bank loan platform by embedding data points in the front-end H5 page, providing the bank with accurate user portraits and risk assessments, and improving the accuracy and efficiency of loan approval.

[0176] The result application module can be used for business optimization:

[0177] Approval process optimization: Analyze the data in the approval process, optimize the approval process, and improve approval efficiency.

[0178] Product and service optimization: Adjust loan products and services based on user needs and behaviors to improve user satisfaction.

[0179] The result application module can be used for monitoring and early warning:

[0180] Abnormal behavior monitoring: Monitor data flow in real time, detect abnormal behavior, and trigger early warning mechanisms in a timely manner.

[0181] Performance indicator monitoring: Monitor key business indicators, such as approval speed, pass rate, etc., to ensure stable business operation.

[0182] The result application module can be used for data visualization:

[0183] Real-time reports: Generate real-time reports and present the analysis results in charts to facilitate management to make decisions quickly.

[0184] Dashboard: Create management dashboards to centrally display key business indicators and monitoring results.

[0185] Decision support: Real-time data: The server processes and analyzes the user's behavior data on the H5 page in real time, such as click behavior, operation frequency, etc. Decision basis: These analysis results serve as an important basis for approval decisions, helping approval personnel or automatic approval systems to quickly make decisions on loan approval or rejection.

[0186] Risk assessment: Risk scoring: Generate a risk score through real-time analysis of user data to reflect the user's credit status and loan risk.

[0187] Approval criteria: Risk scores directly affect approval results. Users with high risk scores may receive faster approval and more favorable loan terms.

[0188] Approval process optimization: Efficiency improvement: Real-time data analysis can help banks identify bottlenecks in the approval process, thereby optimizing the process and improving approval efficiency. Automated approval: For low-risk loan applications, automated approval can be achieved to reduce manual intervention and speed up approval.

[0189] Personalized services: Customized approval: Provide personalized approval suggestions and loan conditions based on the user's risk score and behavior data. Customized marketing: Analyze user behavior, provide users with customized loan products and services, and improve marketing conversion rate.

[0190] Risk control: Risk identification: By analyzing data in real time, the server can promptly identify potential high-risk loan applications and take corresponding risk control measures. Early warning system: Establish an early warning system to monitor the approval status of loan applications in real time and promptly detect and handle risks.

[0191] User experience: Improved transparency: Real-time feedback on approval status and decision-making basis to improve user experience and satisfaction.

[0192] Feedback loop: Collect user feedback and loan performance, continuously optimize approval processes and models, and improve approval quality and user satisfaction.

[0193] Real-time data stream: The server receives data from the H5 page in real time, such as user behavior data, personal information, etc.

[0194] Data analysis engine: Use the big data analysis engine to analyze real-time data (user data collected in real time) to generate risk scores, risk assessments and other results.

[0195] Approval decision support system: Feedback data analysis results to the approval decision support system in real time to assist the approval personnel or automatic approval system in making decisions.

[0196] Approval process optimization: Based on real-time analysis results, optimize the approval process to improve approval efficiency and quality.

[0197] User feedback loop: Collect user feedback and loan performance, continuously optimize approval processes and models, and improve approval quality and user satisfaction.

[0198] The technical solution of the present invention can significantly speed up the approval process. The core advantages of the technical solution of the present invention are automation and efficiency improvement, which are specifically reflected in the following aspects:

[0199] Automated approval: Using real-time analysis data, the server can automatically generate risk scores and risk assessment results, thereby quickly making decisions on loan approval or rejection. The automated approval process reduces manual intervention, making the approval process faster and more efficient.

[0200] Reduce human errors: Manual approval may be affected by factors such as the approver's experience and fatigue, while the automated approval system is based on data and algorithms, which reduces human errors and improves the accuracy of approval.

[0201] Real-time decision support: Data processed and analyzed in real time can provide real-time support for approval decisions, allowing approvers to make decisions based on the latest data rather than outdated information.

[0202] Process optimization: Through real-time monitoring and data analysis, bottlenecks in the approval process can be identified, thereby optimizing the process and improving approval efficiency. For low-risk loan applications, automated approval can be achieved to reduce approval time.

[0203] Personalized approval: Based on user data analysis, personalized approval suggestions and loan conditions can be provided, thereby speeding up the approval process and improving the quality of approval.

[0204] The server processes and analyzes the collected data in real time, including not only loan demand and credit status, but also the following aspects:

[0205] User behavior analysis: Browsing path and dwell time: Analyze the browsing path and dwell time of users on the H5 page to understand the user's interests and needs. Click events: Analyze the user's operating habits and preferences through the user's behavior of clicking different buttons or links. Operation frequency: Count the frequency of users in different operations (such as filling out forms, submitting applications) to evaluate user participation and activity.

[0206] Device information: Device type and operating system: Identify the user's device type and operating system to facilitate personalized services. Geographic location: Understand the user's location information through GPS data for risk assessment and market positioning.

[0207] Personal information: Basic information: Analyze basic information submitted by users, such as age, occupation, etc., to assess loan demand. Financial status: Analyze financial information submitted by users, such as income, liabilities, etc., to assess repayment ability.

[0208] Risk Assessment: Risk Scoring: Use machine learning models to assess user risk and evaluate loan risk.

[0209] Default prediction: Predict the user's future default probability to help banks formulate loan strategies.

[0210] Business optimization: Approval process optimization: Analyze the data in the approval process, optimize the approval process, and improve approval efficiency. Product and service optimization: Adjust loan products and services according to user needs and behaviors to improve user satisfaction.

[0211] Monitoring and early warning: Abnormal behavior monitoring: Real-time monitoring of data flow, detection of abnormal behavior, and timely triggering of early warning mechanisms. Performance indicator monitoring: Monitoring key business indicators, such as approval speed, pass rate, etc., to ensure stable business operation.

[0212] Data visualization: Real-time reports: Generate real-time reports and present the analysis results in charts to facilitate management to make decisions quickly. Dashboard: Create management dashboards to centrally display key business indicators and monitoring results.

[0213] Through real-time processing and analysis of these contents, the technical solution of the present invention can provide banks with in-depth business insights, optimize user experience, improve approval efficiency, reduce operating costs, and effectively control risks.

[0214] Calculation method: 1. Risk score calculation:

[0215] Data collection: Collect multi-dimensional data of users, including but not limited to basic personal information (such as age, occupation, income, etc.), business information (such as business size, revenue, profitability, etc.), credit history (such as past loan records, repayment status, etc.), and behavioral data (such as operating behavior and dwell time on the H5 page).

[0216] Indicator setting: Determine the indicator weights corresponding to each data. For example, credit history may have a higher weight because it can directly reflect the user's repayment ability and credit status. Business information will also have a certain weight to reflect the user's economic strength and stability.

[0217] Scoring model: Use mathematical models to comprehensively calculate various indicators. Common models include linear regression, logistic regression, decision tree, etc. Through these models, the values ​​of different indicators are converted into a specific risk score.

[0218] 2. Loan amount and interest rate adjustment: Loan amount adjustment: Generally speaking, the higher the risk score, the higher the loan amount may be. Different risk score intervals can be set, corresponding to different loan amount ranges. For example, users with high risk scores may get a higher loan amount limit, while users with low risk scores may get a relatively lower loan amount. At the same time, the actual needs and risk tolerance of users are taken into consideration. If the user's business is in good condition and has a clear purpose of funds and repayment plan, the loan amount can be appropriately increased even if the risk score is not particularly high.

[0219] Interest rate adjustment: The interest rate level is also determined based on the risk score. Users with high risk scores have relatively low risks, so the interest rate can be set lower. Users with low risk scores have higher risks, so the interest rate is increased accordingly. Different risk score intervals can be set, corresponding to different interest rates. In addition, adjustments can also be made based on factors such as market interest rates and bank funding costs. For example, when the market interest rate is low, even users with low risk scores may get relatively low interest rates.

[0220] Improve approval accuracy: Multi-dimensional data evaluation: Through data embedding, we can collect various information about users, such as business data (including revenue, profit, tax payment, etc.), personal credit data (such as past loan records, repayment status, credit rating, etc.), and behavioral data (operation behavior on the H5 page, stay time, etc.). Comprehensively integrating these multi-dimensional data to evaluate the user's credit status and repayment ability is more accurate and objective than relying solely on one or two pieces of information. For example, although a company's revenue has declined slightly recently, it has a good repayment record in the past, and it has carefully filled in the information on the H5 page and carefully checked the relevant terms. Combining this information can more accurately assess its risk level, rather than simply rejecting loans or granting lower amounts based on a decline in revenue.

[0221] Dynamic real-time analysis: Real-time data processing can promptly reflect the user's current status and changes. For example, if a company has recently received a large order, or an individual's asset status has improved, these real-time information can be captured and included in the analysis in a timely manner, so as to more accurately adjust the loan amount and interest rate. Compared with the traditional method of regular evaluation or reliance on static data, this dynamic real-time analysis can better adapt to market changes and dynamic fluctuations in user conditions, making the approval results more in line with the user's actual situation and improving accuracy.

[0222] Identify risk signals: In the process of real-time processing and analyzing data, potential risk signals can be quickly identified by setting some risk indicators and early warning mechanisms. For example, if a user frequently inquires about the loan amount or interest rate in a short period of time, it may indicate that the capital demand is urgent or there are other potential problems. Or if a certain financial indicator of the enterprise suddenly deteriorates, the system can detect it in time and conduct in-depth analysis so that more cautious measures can be taken during approval, such as reducing the amount or raising the interest rate, thereby effectively reducing the risk of default and improving the accuracy of approval.

[0223] Improve approval efficiency:

[0224] Automated decision-making process: Automatically adjust loan amounts and interest rates based on data analysis results, reducing manual intervention. In the traditional approval method, loan officers need to spend a lot of time collecting information, analyzing and judging, and then deciding on loan amounts and interest rates. This process is not only inefficient, but also easily affected by human factors, such as loan officers' experience and subjective judgment. After data collection is completed, the automated decision-making process can quickly calculate and adjust according to preset algorithms and rules, greatly shortening the approval time and improving overall efficiency.

[0225] Quick response to market demand: Ability to make real-time adjustments based on market conditions and user needs. For example, when market funds are tight, banks can adjust interest rates in a timely manner through data analysis to attract high-quality customers. Or when a certain industry is developing well, the system can automatically provide more favorable credit and interest rate policies to companies in the industry, quickly meet market demand, improve business competitiveness, and also improve approval efficiency to avoid customer loss due to slow approval.

[0226] Optimize resource allocation: By analyzing a large amount of data, approval resources can be allocated more reasonably. Applications with lower risks and clear quotas and interest rate adjustments can be approved quickly, while more manpower and time can be allocated for in-depth review of complex or higher-risk applications. This will enable the bank's approval resources to be used more effectively and improve overall approval efficiency.

[0227] 1. Business optimization and decision support:

[0228] Product improvement: The number of active users can reflect the attractiveness and user stickiness of the product. If the number of active users is low, the user behavior data can be analyzed to find out the reasons for user churn and improve the product functions, interface design or user experience in a targeted manner. For example, if it is found that the user churn rate is high in a certain operation link, the process of this link can be optimized to make it more concise and smooth. The number of transaction users directly reflects the commercial value of the product. By analyzing the behavioral characteristics, preferences and needs of transaction users, the terms, amount range, interest rate settings, etc. of loan products can be optimized to better meet user needs and improve the market competitiveness of products.

[0229] Adjustment of marketing strategy: The growth trend of the number of registered users can be used to evaluate the effectiveness of marketing activities. If the number of registered users grows slowly, it may be necessary to adjust marketing channels, advertising strategies, or promotional activities. For example, if data analysis shows that the quality of registered users brought by a certain channel is high, investment in this channel can be increased. Changes in the number of active users and the number of transaction users can help determine marketing priorities. If the number of active users is high but the number of transaction users is low, it may be necessary to strengthen conversion marketing for active users, such as launching exclusive offers, providing personalized loan plan recommendations, etc., to improve transaction conversion rates.

[0230] 2. Risk Assessment and Management:

[0231] Credit risk assessment: By combining data such as the number of registered users, the number of active users, and the number of transaction users, a more comprehensive user profile can be established to assist in credit assessment. For example, users who are long-term active and have transaction records may be considered to have relatively low risk, while newly registered users or users with low activity may require more cautious credit review. Analyzing the repayment behavior and overdue status of different types of users can optimize the risk scoring model and improve the accuracy of risk identification. For example, if it is found that the overdue rate of a certain type of active users is low, this type of user can be given a more favorable interest rate or a higher credit limit.

[0232] Fraud risk prevention: By monitoring the abnormal growth of the number of registered users and the sudden changes in user behavior patterns, potential fraud can be discovered in a timely manner. For example, if the number of registered users increases significantly in a short period of time, and there are similarities or abnormal behaviors in the information of these users, it may indicate the risk of batch registration fraud. Changes in the number of transaction users can also serve as an early warning indicator of fraud risk. If a user suddenly conducts transactions frequently or the transaction amount is abnormal, it may be necessary to further review the authenticity and rationality of his transactions.

[0233] 3. Resource Allocation and Planning:

[0234] Human resource arrangement: Based on the fluctuations in the number of active users and transaction users, human resources such as customer service personnel and approval personnel can be reasonably arranged. During business peaks, additional staff can be added to ensure timely processing of user inquiries and loan applications. During business troughs, personnel training or business process optimization can be carried out to improve work efficiency.

[0235] Technical resource optimization: By analyzing data traffic and user behavior, we can optimize server configuration and network bandwidth to ensure system stability and response speed. For example, if we find that user traffic is concentrated in a certain period of time, we can increase server resources in advance to avoid system freezes or crashes.

[0236] According to the changes in user needs and usage habits, timely adjust the direction of technology research and development, and develop functions and services that better meet user needs. For example, if user demand for mobile terminals increases, you can increase the optimization and function expansion of mobile terminal H5 pages.

[0237] 1. Data collection and analysis stage:

[0238] Comprehensively collect user behavior data: Collect various user behaviors on H5 pages through embedding points, including clicks, slides, inputs, dwell time, etc. For example, record the user's dwell time on different pages to understand the user's attention to each link. Collect user device information, network status, etc., in order to analyze external factors that may affect user experience.

[0239] In-depth data analysis: Analyze the user's behavior at key nodes in the application process. For example, check how many users quit while filling out the form. The reasons for quitting may be complex form design, too many required items, etc. Analyze the page jump path to understand whether the user follows the expected process. If a large number of users choose an unexpected path after a certain page, it may mean that the guidance of the page is not clear enough.

[0240] 2. Problem identification and optimization direction determination stage:

[0241] Identify the problem: Based on the data analysis results, identify the specific problems that affect the user experience. For example, if you find that users stay too long on a page and the bounce rate is high, it may be that the page loads slowly or the content is difficult to understand.

[0242] Pay attention to issues that users frequently report, such as complicated application process and unclear information prompts.

[0243] Determine the optimization direction: Slow loading speed: optimize the page resource loading strategy, compress images, reduce unnecessary script loading, etc. For complex form design: simplify the form content, reasonably arrange the required and optional items, and provide clear filling instructions and examples.

[0244] If users are easily lost in the process, optimize page navigation and guidance, using clear button labels and process instructions.

[0245] 3. Optimization implementation and verification phase:

[0246] Implementation of optimization measures:

[0247] According to the determined optimization direction, make corresponding adjustments and improvements to the H5 page. For example, redesign the form layout, increase the page loading speed, optimize the display of prompt information, etc.

[0248] During the implementation process, ensure that the optimization measures will not introduce new problems, such as compatibility issues, functional anomalies, etc.

[0249] Verify the optimization effect: Carry out data collection again to collect user behavior data on the optimized H5 page. Compare the data before and after optimization to evaluate the effect of the optimization measures. If the user stay time is shortened, the bounce rate is reduced, and the application completion rate is increased, it means that the optimization measures are effective. If the effect is not obvious, further analyze the reasons and adjust the optimization plan.

[0250] Data application solutions:

[0251] Loan approval optimization: Application: Use data analysis results (such as risk scores, repayment ability scores) to optimize the loan approval process. Optimization plan: For users with high risk scores, loans can be automatically approved or more favorable loan terms can be provided. For users with low risk scores, loans can be further reviewed or rejected.

[0252] Personalized service provision: Application: Provide personalized loan products and services based on user profiles and preferences. Optimization plan: Customize loan products for different user groups, such as providing low-interest loans for students and flexible repayment plans for small businesses.

[0253] Enhanced risk control: Application: Use data analysis to identify high-risk loans and take preventive measures. Optimization plan: For identified high-risk loans, you can increase guarantee requirements, adjust loan amounts or interest rates, or implement stricter repayment monitoring.

[0254] Marketing strategy adjustment: Application: Analyze user behavior and needs to optimize marketing strategies. Optimization plan: Develop targeted marketing activities based on user characteristics and behaviors to improve marketing conversion rates.

[0255] User experience improvement: Application: Improve user experience by analyzing user behavior on H5 pages. Optimization plan: Simplify the application process, optimize page layout, provide useful guidance and help, and reduce user churn.

[0256] Specific optimization plan:

[0257] Automation of approval process: Access the bank's automatic approval system to automatically make approval decisions based on data analysis results.

[0258] Dynamic risk pricing: Dynamically adjust loan interest rates and limits based on risk scores to balance risks and returns.

[0259] User behavior feedback loop: Establish a feedback mechanism to incorporate users’ repayment behavior and loan performance into data analysis and continuously optimize the model.

[0260] Data-driven marketing: Use data analysis results to implement precision marketing and improve marketing efficiency.

[0261] User interface optimization: Based on user behavior data, optimize the design and functionality of the H5 page to improve user satisfaction and conversion rate.

[0262] Monitoring and early warning system: Implement real-time monitoring and early warning system to promptly detect and handle potential risks.

[0263] Continuous learning and model updating: Data analysis models are regularly evaluated and updated to adapt to market changes and changes in user behavior.

[0264] Through these optimization schemes, the technical solution of the present invention can more effectively utilize data analysis results, improve the efficiency of business processes, reduce operating costs, and at the same time improve risk management capabilities and customer service quality.

[0265] The specific risk assessment program includes the following steps:

[0266] Data collection: When users submit loan applications through the H5 page, the user's behavioral data and application information are collected.

[0267] Feature extraction: Extracting features from the collected data for risk assessment.

[0268] Model application: These features are input into the trained risk assessment model.

[0269] Score calculation: The model calculates a risk score that takes into account factors such as the user’s credit history, financial status, behavioral patterns, etc.

[0270] Approval decision: Make loan approval decisions automatically or manually based on risk scores and preset approval criteria.

[0271] Feedback loop: Approval results and subsequent repayment behavior are fed back to the model for continuous optimization of the model.

[0272] In this way, loan risks can be effectively assessed, approval efficiency and accuracy can be improved, and default risks can be reduced.

[0273] With the rapid development of Internet technology, financial services have gradually shifted from offline to online, and online loan products such as bank loan platforms have become important financing tools for enterprises. However, the effective collection, analysis and utilization of full-process data of online loan products is the key to improving product service quality, optimizing user experience and reducing risks. At present, data embedding technology in the front-end H5 page has become an important means of collecting user behavior data, but its application in the financial field still has certain limitations and shortcomings.

[0274] The present invention provides an information processing system based on the front-end H5 full-process data embedding of the bank loan platform. Through the refined data embedding design, it can accurately capture and deeply analyze the full-process behavior of users, and provide strong support for product optimization, user experience improvement and risk control. The implementation of the present invention can significantly improve the service quality and user experience of online loan products, reduce risks, and has broad application prospects and social value.

[0275] The purpose of the present invention is to collect and integrate user behavior data and external user individual label data to supplement it, optimize the product, analyze the data and customize better operation strategies, combine multi-dimensional information when problems arise, correctly iterate the functions, maximize the product effect, improve marketing effects, and enhance risk control capabilities.

[0276] With the rapid development of mobile Internet, more and more users choose to conduct financial business through mobile phones. As an online loan product, the bank loan platform needs to collect user behavior data in real time during the application process in order to improve user experience and approval efficiency, so as to conduct risk assessment and build user portraits. However, traditional data collection methods often have problems such as data loss and delay, and cannot meet the bank loan platform's requirements for data real-time and accuracy.

[0277] The present invention aims to provide an information processing system based on the front-end H5 full-process data embedding of the bank loan platform. Through refined data embedding design, it can realize the accurate capture and in-depth analysis of the user's full-process behavior in the front-end H5 page of the bank loan platform, and provide strong support for product optimization, user experience improvement and risk control.

[0278] Objectives achieved:

[0279] (1) Construction of the foundation of global user data. Complete the collection of user behavior through all channels, including the self-operated entrance of WeChat public accounts, the QR code entry of H5 marketing partner organizations, the industrial bank APP, the unified entry system, and the collection of business process nodes of the comprehensive credit management system, and integrate the external individual tags of corporate legal persons / shareholders / executives.

[0280] (2) User experience optimization. Optimize core user paths such as real-name authentication, credit application, credit application, and repayment application, analyze the reasons for process failure, reduce user friction, reduce user churn and customer complaints, and shorten user time.

[0281] (3) Improve the effectiveness of marketing operations. With the help of external user data, accurately identify the characteristics of the target audience and the activity of similar lending platforms. With the help of full-link channel interaction data analysis, optimize channel delivery strategies, reduce customer acquisition costs, and improve customer acquisition efficiency. Combine user activity, credit usage / repayment status, channel source, etc. to conduct tiered operations on users, and accurately recall users who have failed in the process, have been granted credit but have not withdrawn funds, have failed to use credit, have good repayment status and have a credit balance. Supplement the labels of old users, including recent activity of similar banks / lending platforms, income status, asset status, family member status, normal use of mobile phones, etc., and launch multiple waves of personalized marketing awakening.

[0282] (4) Improved risk control capabilities. Identify abnormal user operation behaviors and frequencies through tracking. Supplement with external user data tags to enrich risk control scenarios, including normal user behavior judgment, such as regular social interaction, news browsing, car maintenance, live video, accounting, map navigation, etc. Judgment of high-risk users, such as high activity of blockchain and virtual currencies, changes in income levels, vehicle transactions, etc. Judgment of multiple loan risks, such as high activity of multiple installment loan products of the same type.

[0283] To achieve the above object, the present invention adopts the following technical solutions:

[0284] Data embedding code is embedded in the front-end H5 page of the bank loan platform to realize real-time collection of user behavior. The data embedding code can collect user behavior data such as clicks, inputs, and slides on the page, and send it to the server for processing in real time.

[0285] Design an indicator system for data collection, including user personal information, behavioral data, device information, etc., to ensure the comprehensiveness and accuracy of the data.

[0286] The collected data is processed and analyzed in real time through the server to build a user profile and risk assessment model. The server can analyze the user's loan needs, credit status and other information based on the user's behavior data to provide a basis for loan approval.

[0287] Based on the data analysis results, the approval process and strategy of the bank loan platform are optimized to improve the accuracy and efficiency of approval. For example, the loan amount and interest rate can be automatically adjusted according to the user's risk score.

[0288] The module functions can be realized through executable programs and code development. The relevant product function modules and functions are shown in Table 1 below.

[0289] Table 1 Related product function modules and function introduction

[0290]

[0291]

[0292]

[0293]

[0294] According to the data analysis results, the approval process and strategy of the bank loan platform are optimized to improve the accuracy and efficiency of approval. For example, the loan amount and interest rate can be automatically adjusted according to the user's risk score. In this way, personalized services and risk control can be achieved for users, and the accuracy and efficiency of loan approval can be improved.

[0295] These include: Basic business improvement: The main analysis indicators include: number of active users, number of newly registered users, retention rate, number of logins, number of logged-in users, login duration, number of incoming applications, credit amount, credit amount, repayment amount, etc. The main analysis dimensions include: various channels (cooperating institutions) / operating system / province / region / application version / function type / user level / product category / time (day / week / month), etc. The analysis results can be visualized, such as Figure 3 The visual analysis results are shown.

[0296] Marketing evaluation improvements: The main analysis indicators include: promotion landing page PV, landing page visitors (UV), landing page bounce rate, landing page element / position click / browsing depth line, new channel user registration number, registration conversion rate, new device retention rate, etc. The main analysis dimensions include: promotion channel / promotion. Keywords / promotion type / active city / network type / browser type / operating system / time (day / week / month) / high activity of similar lending platforms and other analysis results.

[0297] Functional improvement: The main analysis indicators include: distribution of SMS verification functions, FaceID verification times, number of FaceID verification users, number of ID uploads, number of ID upload users, number of scan times, number of scan users, average number of times a function is used, retention rate of a function, page dwell time, bounce rate, step conversion time, etc. The main analysis dimensions include: various channels (cooperating institutions) / operating systems / provinces / regions / application versions / function types / user levels / product categories / time (day / week / month), etc.

[0298] Improvement of business operations: The main analysis indicators include: number of clicks on the operation position, number of clickers on the operation position, number of clicks on the operation position per capita, number of clicks on different operation positions, number of clickers on different operation positions, number of clicks on different operation positions per capita, user usage path, conversion rate of personal credit application process, conversion rate of corporate credit application process, number of credit applications, number of users who applied for credit, number of successful credit submissions, number of users who successfully submitted credit, step conversion time, conversion rate of personal credit application process, conversion rate of corporate credit application process, number of credit applications, number of users who applied for credit, number of successful credit submissions, number of users who successfully submitted credit (number of users withdrawing funds), number of failed credit submissions, number of users who failed to submit credit, distribution of reasons for failed credit submissions, number of repayment applications, number of users who applied for repayment, number of successful repayment submissions, number of users who successfully submitted repayments, number of failed repayment submissions, etc. The main analysis dimensions include: various channels (cooperating institutions) / operating systems / provinces / regions / application versions / function types / / product categories / time (day / week / month), etc.

[0299] The innovations of this patent are mainly reflected in the following aspects:

[0300] Front-end H5 full-process data embedding technology:

[0301] By embedding data embedding code in the front-end H5 page, the real-time collection of user behavior data in the application process of the bank loan platform is realized. This technology can capture user clicks, inputs, sliding and other behaviors, providing banks with detailed data on user interactions.

[0302] Construction of data collection index system: A comprehensive data collection index system has been designed, including user personal information, behavior data, device information, etc., to ensure the comprehensiveness and accuracy of the data. Such an index system helps to more accurately analyze user behavior and credit risk.

[0303] Real-time data processing and analysis: The collected data is processed and analyzed in real time through the server to build user portraits and risk assessment models. This real-time processing capability enables banks to quickly respond to market changes and user needs and improve decision-making efficiency.

[0304] Approval process and strategy optimization: Based on the data analysis results, the approval process and strategy of the bank loan platform are optimized to improve the accuracy and efficiency of approval. For example, the loan amount and interest rate can be automatically adjusted according to the user's risk score to achieve personalized services and risk control.

[0305] Digital transformation and improved user experience: This patented technology helps promote the digital transformation of banking business, improves user experience and service quality through real-time collection and analysis of user data, and enhances the competitiveness of banks.

[0306] Broad application prospects: The technical solution of the present invention is not only applicable to products of bank loan platforms, but can also be widely used in other financial business scenarios and various banking business scenarios, and has broad market prospects and application value.

[0307] The advantages of the present invention are: real-time collection of user behavior data, ensuring the timeliness and accuracy of the data. A comprehensive data indicator system is constructed to improve the accuracy of user portraits and risk assessments. The approval process and strategy are optimized to improve the efficiency and accuracy of loan approval. It can be widely used in various financial business scenarios and various banking business scenarios, and has broad market prospects and application value.

[0308] The information processing system in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0309] (1) The technical solution of the present invention solves the following technical problems through the front-end H5 full-process data embedding solution:

[0310] Process efficiency and automation: The bank has achieved rapid processing of the entire process from customer application to bank approval, with the fastest time being only 7 minutes. This efficiency improvement is due to the application of front-end data embedding technology, which makes the entire process highly automated and reduces manual intervention.

[0311] Data-driven and decision-making support: Through data embedding technology, banks can collect and analyze customer behavior data, and then use big data models to support decision-making. For example, banks can understand the operating rate of customer equipment through the GPS system and analyze it in combination with other data to decide whether to issue a loan.

[0312] User experience optimization: Data embedding technology helps banks optimize customer experience. Customers can complete all processes through their mobile phones, including identity authentication, information entry, loan application and approval, etc. This convenience greatly improves user satisfaction.

[0313] Risk Control and Business Management: Patented technology also helps banks make progress in risk control and business management. Through real-time monitoring and analysis of data, banks can better assess and manage loan risks while improving the speed and accuracy of business processing.

[0314] In traditional financial business processes, especially loan application and approval processes, the following methods are usually used:

[0315] Paper document processing: Traditional methods usually involve a large amount of paper documents, including loan application forms, identity certificates, financial statements, etc. These documents need to be filled out manually, signed, and submitted to the bank by mail or in person.

[0316] Manual review: Bank staff manually review the documents submitted by applicants to check their completeness and accuracy. This process is usually time-consuming and prone to errors.

[0317] Face-to-face service: In the traditional process, customers may need to go to a bank branch in person for consultation, submit application materials, and complete other related procedures.

[0318] Communication by phone or email: During the application process, communication between customers and banks is mainly conducted by phone or email, which may result in untimely information delivery or inefficient communication.

[0319] Simple data collection: While some traditional systems may include basic electronic data collection capabilities, these systems often lack in-depth analysis of user behavior and comprehensive monitoring of the entire application process.

[0320] Independent system operation: Different business links may be processed using different systems, such as application entry system, approval system, post-loan management system, etc. There may be a lack of effective data exchange and integration between these systems.

[0321] Compared with these traditional methods, the patented front-end H5 full-process data embedding technology of the technical solution of the present invention provides a more efficient, automated, and data-driven processing method. Through this technology, banks can achieve rapid approval, precision marketing, optimize user experience, and improve overall business processing capabilities.

[0322] (2) The drawbacks and shortcomings of traditional technologies in handling loan application and approval processes mainly include:

[0323] Inefficiency: Paper documents and manual processing processes are often very time-consuming. Each application needs to be manually reviewed, which can lead to slow processing speeds and increase customer waiting time. High error rate: Manual processing is prone to errors, whether it is data entry errors or omissions during the document review process, which may cause process delays or wrong decisions. High cost: Maintaining paper documents, storage space, human resources, etc. requires high costs. In addition, processing errors and rework will also add additional costs. Poor user experience: Customers need to visit the bank in person or communicate through emails and phone calls. This process can be cumbersome and inconvenient, affecting the overall customer experience. Limited data analysis and decision support: Traditional methods often lack in-depth analysis and mining of data, which limits banks' ability to use data to optimize products and services and improve decision-making quality. Information islands: There is often a lack of effective data sharing and integration between different systems, resulting in information islands and inefficient data utilization. Security issues: Paper documents are easily lost, damaged or tampered with, and traditional data storage systems may lack sufficient security measures to protect sensitive information. Lack of transparency: Each link in the traditional process may not be transparent enough, and it is difficult for customers to track the status of applications, which may lead to a decline in trust. Poor adaptability and scalability: With the increase in business volume or market changes, traditional systems may find it difficult to quickly adapt to new needs and have poor scalability. Environmental impact: The extensive use of paper documents has a negative impact on the environment and does not meet the requirements of sustainable development. Therefore, in order to overcome these drawbacks and shortcomings, financial institutions are increasingly adopting digital and automated solutions, such as front-end H5 full-process data embedding technology, to improve efficiency, reduce costs, optimize user experience, and enhance data analysis and decision support capabilities.

[0324] (3) Solve the drawbacks and shortcomings of traditional technologies by:

[0325] Digital processing of the entire process: By transferring the entire loan application process to the H5 platform, the entire process from application to approval is digitized, reducing the use of paper documents and improving processing speed.

[0326] Data collection and real-time monitoring: Implementing data collection on H5 pages can collect user behavior data in real time, providing banks with instant business insights, thereby optimizing processes and improving user experience.

[0327] Automated approval process: Utilize the collected data and combine it with machine learning algorithms to automate loan approval, reduce manual intervention, lower the error rate, and improve approval efficiency.

[0328] User-friendly interface design: H5 technology supports user-friendly interface design, making the loan application process more intuitive and convenient, and improving the user's operating experience.

[0329] Data integration and analysis: By integrating data from different systems, information silos are broken, allowing banks to conduct more comprehensive data analysis and risk assessment.

[0330] Enhanced security: Advanced encryption technology and security protocols are used to ensure the security of user data transmission and storage on the H5 platform.

[0331] Cost savings: Reduced costs for paper document processing, storage, and manual review, and further reduced operating costs through automated processes.

[0332] Environmentally friendly: It reduces the reliance on paper documents, helps protect the environment, and complies with the concept of sustainable development.

[0333] Flexibility and scalability: The flexibility of the H5 platform enables banks to quickly adjust and expand services based on market demand, improving the adaptability of their business.

[0334] (4) Specifically, the following are several key points of the technical solution of the present invention to solve the drawbacks and shortcomings:

[0335] Fast approval: Through the real-time data collection and automated approval of the H5 platform, the loan approval time is greatly shortened from the traditional days or even weeks to as fast as 7 minutes.

[0336] Decision support: Using big data analysis and machine learning to make loan decisions based on user behavior data, improving the accuracy and efficiency of decision-making.

[0337] User experience: Users can complete the entire loan process on their mobile phones without going to the bank, which simplifies the operation steps and improves satisfaction.

[0338] Risk control: By real-time monitoring and analyzing data, banks can better manage loan risks and ensure business stability.

[0339] (5) The specific analysis method of the technical solution of the present invention for real-time processing and analysis of the collected data by the data processing and analysis module on the back-end server may include the following aspects:

[0340] User behavior analysis: 1) Clickstream analysis: Track user click behavior and analyze user interaction patterns on H5 pages, such as which buttons are clicked and which pages are visited. 2) Path analysis: Analyze the user's path from entering the application to completing the loan application, and identify common user behavior paths and potential loss points.

[0341] Data mining: 1) Feature extraction: Extract key features from the collected data, such as user personal information, device information, behavior patterns, etc. 2) Pattern recognition: Use machine learning algorithms to identify patterns in user behavior, such as which behaviors are related to loan approval rates.

[0342] Machine Learning Models: 1) Risk Scoring Model: Using machine learning techniques such as random forests and gradient boosting machines, a risk scoring model is built based on user data for automatic approval processes. 2) Risk Prediction Model: Predict loan default risks to help banks decide whether to approve loan applications.

[0343] Real-time monitoring and early warning: 1) Anomaly detection: Real-time monitoring of data streams, using anomaly detection algorithms to detect abnormal behaviors and trigger early warning mechanisms in a timely manner. 2) Performance indicator monitoring: Monitor performance indicators of the loan approval process, such as approval speed, pass rate, etc.

[0344] Statistical analysis: 1) Conversion rate analysis: Calculate the conversion rate of key links such as visit to application, application to approval, etc. 2) A / B test analysis: Test different page layouts or process designs to analyze which method is more effective.

[0345] Big data analysis: 2) User grouping: Divide users into different groups based on their behavioral characteristics, credit status, etc. to achieve precision marketing. 2) Trend analysis: Analyze market trends and changes in user needs to provide data support for product iteration and market strategy.

[0346] Data visualization: 2) Real-time reports: Generate real-time reports and present the analysis results in charts to facilitate management to make decisions quickly. 3) Dashboard: Create management dashboards to centrally display key business indicators and monitoring results.

[0347] These analytical methods combined can provide in-depth business insights for bank operations, optimize user experience, improve approval efficiency, reduce operating costs, and effectively control risks.

[0348] (6) The process of building a user profile and risk assessment model is roughly as follows:

[0349] Build user portraits: 1) Data collection: Collect basic information about users, such as age, gender, occupation, etc. Obtain user transaction behavior data, browsing behavior, click data, etc. 2) Feature engineering: Extract useful features from the collected data, such as consumption habits, repayment ability, social activities, etc. Clean, transform and normalize the features. 3) User grouping: Use clustering algorithms (such as K-means) to group users and find groups with similar characteristics. Define labels for each group, such as "high-income group", "student group", etc. 4) Portrait construction: Build a detailed portrait of each user based on the features and grouping results. User portraits may include multiple dimensions, such as demographic characteristics, consumption characteristics, credit characteristics, etc.

[0350] Building a risk assessment model: 1) Data preparation: Collect historical loan data, including loan amount, repayment status, overdue records, etc. Combined with user portrait data, prepare a complete data set for modeling. 2) Model selection: Select an appropriate machine learning algorithm, such as logistic regression, decision tree, random forest, etc. 3) Feature selection: Select the features that have the greatest impact on risk assessment through feature importance analysis. 4) Model training: Use the training data set to train the model and adjust the model parameters to achieve optimal performance. Evaluate the accuracy and generalization ability of the model through methods such as cross-validation. 5) Model validation: Use the validation data set to test the model to ensure the effectiveness of the model in practical applications.

[0351] Subsequent use: 1) Use of user portraits: Personalized services: Provide personalized loan products and services based on user portraits. Marketing strategies: Develop precise marketing strategies based on the characteristics of different user groups. User experience optimization: Improve the design and functions of the H5 page based on user portraits to enhance user experience. 2) Use of risk assessment models: Automatic approval: When a user submits a loan application, the risk assessment model is automatically used for approval to improve approval efficiency. Risk control: Identify high-risk loan applications and take measures to reduce potential default risks. Decision support: Provide data support to bank managers to help make more informed loan decisions. In this way, banks can manage loan business more effectively, improve service quality, and reduce operating costs and risks. The server can analyze user loan needs, credit status and other information based on user behavior data.

[0352] (7) Scoring Type: In the technical solution of the present invention, the scoring may include the following types: Loan demand score: This score reflects the user's demand for loans, which may be based on factors such as the user's behavior pattern, application information, historical loan records, etc. Credit status score: This score assesses the user's credit risk, usually based on the user's repayment record, credit history, financial status, etc.

[0353] In addition to the above scores, there may be the following scores: Repayment Ability Score: Assess the user's ability to repay the loan, which may take into account the user's income level, assets and liabilities, etc. Fraud Risk Score: Assess the authenticity of the information submitted by the user, used to detect potential fraud. Comprehensive Score: Combines the above multiple scoring dimensions to provide users with a comprehensive evaluation result.

[0354] These scoring results can be used in the following ways:Automatic approval: Automatically decide whether to approve a loan application based on the scoring results.Risk pricing: Adjust the loan interest rate or amount based on the risk score.Personalized service: Provide customized loan products based on the loan demand score.

[0355] (8) The impact of the method of providing the basis for loan approval on the approval process is mainly reflected in the following aspects:

[0356] 1) Improved approval efficiency: Automated processing: By collecting and analyzing user behavior data in real time, the credit status and repayment ability of loan users can be automatically evaluated. The automated scoring model can quickly give approval suggestions, reducing the time for manual review, thereby significantly improving approval efficiency.

[0357] 2) Enhanced approval accuracy: Data-driven decision-making: Using big data and machine learning technology, the approval process no longer relies solely on manual judgment, but is analyzed based on large amounts of data and complex algorithms. This data-driven decision-making method can more accurately identify potential high-quality customers and high-risk customers.

[0358] 3) Risk control optimization: refined management: Data embedding technology enables banks to collect more dimensional information, such as users' browsing behavior, operating habits, etc., which helps to assess risks more finely. By analyzing this data, banks can better identify and prevent fraud and reduce the risk of loan defaults.

[0359] 4) Approval process transparency: Process monitoring: Patented technology allows banks to monitor every step of the approval process in real time to ensure the transparency and fairness of the process. Management can view the approval status and decision-making basis at any time, making it easier to monitor and adjust approval strategies.

[0360] 5) Specific ways of impact: Preliminary screening: After the user submits the application, the system will immediately conduct a preliminary screening based on the preset rules and models to quickly exclude applications that do not meet the requirements. Risk scoring: The user's risk score directly affects the approval result. Users with high risk scores may receive faster approval and more favorable loan terms. Behavioral analysis: Analyze the user's behavior patterns on the H5 page, such as whether they repeatedly check the loan conditions, whether they are hesitant, etc. These behavioral characteristics can be used as reference factors for approval. Dynamic adjustment: The approval model can dynamically adjust the approval standards according to market conditions and bank strategies to ensure the flexibility and adaptability of the approval process. Instant feedback: The system can provide users with instant approval feedback, and can provide improvement suggestions or alternatives for rejected applications.

[0361] (9) Loan amount adjustment method: Data collection and analysis: Collect users' basic information, financial status, historical credit records, behavioral data, etc. Analyze this data to evaluate users' repayment ability, credit risk, etc. Risk scoring model: Use machine learning algorithms to train risk scoring models based on user data. The risk scoring model outputs a score that reflects the user's credit rating. Quota rule setting: Determine the loan amount range corresponding to different credit ratings based on risk scores and preset rules. The rules may include factors such as credit score and quota mapping table and risk tolerance. Dynamic adjustment: Dynamically adjust the quota rules based on market conditions, bank risk preferences and policy orientations. For specific user groups or market activities, temporary quota adjustment strategies can be implemented. User feedback and adjustment: Adjust the loan amount of individual users based on user feedback and loan performance.

[0362] (10) Interest rate adjustment methods: Risk pricing model: Establish a risk pricing model that takes into account factors such as the user's credit risk, market interest rates, and bank costs. The model will calculate a risk-based loan interest rate. Interest rate rule setting: Set interest rate rules to correspond risk scores to interest rate levels. Rules may include interest rate ranges for different credit ratings, fixed markups, or floating ratios. Market interest rate linkage: Consider changes in market interest rates, such as changes in the benchmark interest rate, and adjust the loan interest rate accordingly. Personalized pricing: Provide personalized interest rate pricing based on factors such as the user's risk score, loan history, and relationship with the bank. Promotions and offers: Provide promotional interest rates or preferential conditions during specific periods or for specific user groups.

[0363] Specific adjustment process: Application submission: Users submit loan applications through the H5 page. Data evaluation: The server collects and analyzes user data, including behavioral data, risk scores, etc. Amount and interest rate determination: Based on risk scores and preset rules, the system automatically determines the loan amount and interest rate. Approval feedback: Feedback the approval results (including amount and interest rate) to the user. User confirmation: The user accepts or rejects the loan conditions provided.

[0364] In this way, the technical solution of the present invention can achieve dynamic and personalized adjustment of loan amount and interest rate to adapt to the risk status and market environment of different users, while optimizing the bank's loan business efficiency and risk management.

[0365] It should be noted that the contents not described in detail in the specification of the present invention belong to the common knowledge of those skilled in the art.

[0366] Figure 4 It is a structural diagram of an information processing method based on a bank loan platform according to an embodiment of the present invention.

[0367] like Figure 4 As shown, an information processing method based on a bank loan platform is provided in an embodiment of the present invention, and the information processing method based on the bank loan platform includes: S101: determining a data point-burying scheme according to the business process and user behavior characteristics of the front-end H5 page of the bank loan platform. The business process includes a loan application process. S102: collecting user data in real time according to the data point-burying scheme, and sending the user data to the back-end server. User data includes user behavior data and application information. Behavior data includes: user browsing path, dwell time and click events. Application information includes: application time, loan amount, loan term, repayment method, personal information: user's name, age and gender, business information: business size and business information, credit history: historical loan records and historical repayment records. S103: receiving and processing and analyzing user data to determine data analysis results. S104: optimizing the loan approval process according to the data analysis results includes: adjusting the loan amount and interest rate, making loan approval decisions, so as to speed up the loan approval process. For details, please refer to the specific implementation method of the above system, which will not be repeated here.

[0368] See also Figure 5 , Figure 5 Schematic diagram of the hardware structure of the computer device according to the embodiment of the present invention. Figure 5 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including a high-speed interface and a low-speed interface. Figure 5 A processor 10 is taken as an example. The processor 10 may be a central processing unit, a network processor or a combination thereof. The memory 20 stores instructions executable by at least one processor 10 so that at least one processor 10 executes the method shown in the above embodiment. The computer device also includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0369] The embodiment of the present invention also provides a computer-readable storage medium, and the method according to the embodiment of the present invention described above can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or downloaded through a network and originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Implement the method shown in the above embodiment.

[0370] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0371] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. An information processing system based on a bank loan platform, characterized in that: include: The data embedding design module is used to determine the data embedding plan based on the business process and user behavior characteristics of the front-end H5 page of the bank loan platform; The business process includes a loan application process; The data collection module is configured in the front-end H5 page, and is used to collect user data in real time according to the data embedding scheme determined by the data embedding design module, and send the user data to the back-end server; The user data includes user behavior data and application information; The behavior data includes: the user's browsing path, dwell time and click events; the application information includes: application time, loan amount, loan term, repayment method, personal information: user's name, age and gender, business information: business size and business information, credit history: historical loan records and historical repayment records; A data processing and analysis module, configured on the back-end server, for receiving, processing and analyzing the user data to determine a data analysis result; The result application module includes a loan approval optimization unit, which is used to optimize the loan approval process according to the data analysis results, including: adjusting the loan amount and interest rate, making loan approval decisions, so as to speed up the loan approval process.

2. The system according to claim 1, characterized in that The data embedding design module includes: A tracking point demand determination unit is used to determine the tracking point demand according to the business process and user behavior characteristics of the front-end H5 page of the bank loan platform; the tracking point demand includes key nodes and data fields corresponding to the key nodes; The tracking point scheme design unit is used to determine the tracking point location according to the key nodes in the tracking point requirements, determine the data type according to the data fields corresponding to the key nodes, and design the corresponding collection frequency.

3. The system according to claim 1, characterized in that The data acquisition module comprises: A data collection unit, used to collect user data in real time according to the data embedding scheme determined by the data embedding design module; A data encryption unit, used to encrypt the user data to obtain encrypted user data; A data compression unit, used for compressing the encrypted user data to obtain compressed user data; The data transmission unit is used to send the compressed user data to the back-end server.

4. The system according to claim 1, characterized in that The data processing and analysis module comprises: A data processing unit, used for receiving the user data and performing data processing on the user data to obtain the processed user data; the data processing includes: preprocessing and feature extraction; The data analysis unit is used to analyze the user data after data processing to obtain the data analysis result.

5. The system according to claim 4, characterized in that The data analysis unit comprises: A model building unit, used to build a risk assessment model; The model application unit is used to input the processed user data as input data into the risk assessment model to obtain the risk score corresponding to the user; the risk score is used to characterize the credit risk and loan risk of the user.

6. The system according to claim 5, characterized in that The loan approval optimization unit is specifically used to adjust the loan amount and interest rate according to the risk score corresponding to the user, and automatically or manually make a loan approval decision according to the risk score corresponding to the user and the preset approval criteria.

7. The system according to claim 5, characterized in that The data analysis unit further includes: The user portrait building unit is used to analyze the user data after data processing and build a user portrait.

8. An information processing method based on a bank loan platform, characterized in that: include: Determine the data embedding plan based on the business process and user behavior characteristics of the front-end H5 page of the bank loan platform; The business process includes a loan application process; Collect user data in real time according to the data embedding solution, and send the user data to the back-end server; The user data includes user behavior data and application information; The behavior data includes: the user's browsing path, dwell time and click events; the application information includes: application time, loan amount, loan term, repayment method, personal information: user's name, age and gender, business information: business size and business information, credit history: historical loan records and historical repayment records; Receive, process and analyze the user data to determine data analysis results; Optimizing the loan approval process based on the data analysis results includes: adjusting the loan amount and interest rate, and making loan approval decisions to speed up the loan approval process.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the information processing method based on the bank loan platform as described in claim 8 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the information processing method based on a bank loan platform as described in claim 8.

Images