Business risk early warning method and device and computer readable storage medium

Abstract

The invention provides a business risk early warning method and device and a computer readable storage medium. The method comprises the following steps: acquiring employee basic information of a target employee and learning behavior information and learning result information of the target employee in a target business knowledge point type; inputting the employee basic information, the learning behavior information of the target employee in the target business knowledge point type and the learning result information into a trained extreme gradient boosting tree XGBOOST model to obtain a performance level prediction value of the target employee in the target business knowledge point type; and according to the performance level prediction value of the target employee in the target business knowledge point type, determining whether the target employee has a business risk. According to the method, the device and the computer readable storage medium, the problem that the business risk of employees caused by insufficient business knowledge learning cannot be timely and effectively found due to lack of a business knowledge learning-based business risk early warning method in the prior art can be solved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a business risk early warning method, apparatus, and computer-readable storage medium. Background Technology

[0002] As companies increasingly emphasize continuous learning for their employees, more and more are building internal learning platforms to enhance their professional skills and business acumen. However, most of these platforms are merely used to assess whether employees have completed learning tasks, without effectively analyzing the learning process and outcomes. In business operations, there is a wealth of knowledge in specific vertical fields and subjectively defined concepts. If employees lack access to this type of knowledge, it can lead to significant business risks when conducting marketing or production activities, such as customer churn, data security breaches, and integrity risks.

[0003] However, existing technologies lack a business risk early warning method based on business knowledge learning, making it impossible to promptly and effectively detect business risks caused by employees' insufficient business knowledge learning. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to address the above-mentioned shortcomings of the prior art by providing a business risk early warning method, device and computer-readable storage medium, so as to solve the problem that the prior art lacks a business risk early warning method based on business knowledge learning, which makes it impossible to timely and effectively detect business risks caused by employees' insufficient business knowledge learning.

[0005] In a first aspect, the present invention provides a business risk early warning method, comprising:

[0006] Obtain basic employee information of the target employee, as well as the target employee's learning behavior and learning outcome information in the target business knowledge point type;

[0007] The employee's basic information, the target employee's learning behavior information and learning result information in the target business knowledge point type are input into the trained extreme gradient boosting tree XGBOOST model to obtain the predicted performance level of the target employee in the target business knowledge point type. The trained XGBOOST model is trained based on the employee's basic information, historical learning behavior information, historical learning result information and historical performance level of multiple employees in the target business knowledge point type.

[0008] Based on the predicted performance level of the target employee in the target business knowledge point type, determine whether the target employee poses a business risk.

[0009] Furthermore, before obtaining the target employee's basic information and the target employee's learning behavior and learning outcome information in the target business knowledge point type, the method further includes:

[0010] Collect uncategorized learning content data and question content data of the target employees in the preset learning platform;

[0011] The unclassified learning content data and the question content data are classified using a trained textCNN model to identify learning content data and question content data under different business knowledge point types.

[0012] Based on the learning content data and question content data under the different business knowledge point types, the learning behavior information and learning result information of the target employee in each business knowledge point type are determined, wherein the different business knowledge point types include the target business knowledge point type.

[0013] Furthermore, before classifying the unclassified learning content data and the question content data using the trained textCNN model to identify learning content data and question content data under different business knowledge point types, the method further includes:

[0014] Acquire learning content data and question content data that have been tagged with business knowledge point types in the learning platform;

[0015] Based on the learning content data and question content data of the labeled business knowledge point types, a textCNN model is constructed and trained to obtain the trained textCNN model.

[0016] Furthermore, the learning behavior information includes: learning start time, article word count, article dwell time, quiz start time, quiz duration, number of quiz retries, and average dwell time per article word count; the learning result information includes: quiz score.

[0017] Furthermore, before inputting the employee's basic information, the target employee's learning behavior information and learning result information into the trained extreme gradient boosting tree XGBOOST model to obtain the predicted performance level of the target employee in the target business knowledge point type, the method further includes:

[0018] The employee basic information of the multiple employees, the historical learning behavior information and the historical learning result information in the target business knowledge point type are used as the feature dataset;

[0019] The historical performance level of the multiple employees in the target business knowledge point type is used as the target variable corresponding to the feature dataset;

[0020] Based on the feature dataset and the corresponding target variables, the XGBOOST model is constructed and trained to obtain the trained XGBOOST model.

[0021] Furthermore, the performance level is categorized as excellent, average, and poor. The determination of whether the target employee poses a business risk based on the predicted performance level of the target employee in the target business knowledge point type specifically includes:

[0022] If the predicted performance level of the target employee for the target business knowledge point type is poor, then the target employee is judged to have business risk.

[0023] Furthermore, if the target employee poses a business risk, the method further includes:

[0024] Relevant measures were taken to provide guidance to the target employees.

[0025] Secondly, the present invention provides a business risk early warning device, comprising:

[0026] The first acquisition module is used to acquire the basic employee information of the target employee, as well as the learning behavior information and learning result information of the target employee in the target business knowledge point type.

[0027] The prediction module, connected to the first acquisition module, is used to input the employee's basic information, the target employee's learning behavior information and learning result information in the target business knowledge point type into the trained extreme gradient boosting tree XGBOOST model to obtain the predicted value of the target employee's performance level in the target business knowledge point type. The trained XGBOOST model is trained based on the employee's basic information, historical learning behavior information, historical learning result information and historical performance level of multiple employees in the target business knowledge point type.

[0028] The early warning module, connected to the prediction module, is used to determine whether the target employee has any business risks based on the predicted performance level of the target employee in the target business knowledge point type.

[0029] Thirdly, the present invention provides a business risk warning device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the business risk warning method described in the first aspect above.

[0030] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the business risk warning method described in the first aspect.

[0031] The business risk early warning method, apparatus, and computer-readable storage medium provided by this invention first acquire the basic employee information of the target employee, as well as the target employee's learning behavior information and learning result information in the target business knowledge point type; then, the basic employee information, the target employee's learning behavior information, and the learning result information are input into a trained extreme gradient boosting tree (XGBOOST) model to obtain a predicted performance level of the target employee in the target business knowledge point type. The trained XGBOOST model is obtained based on the basic employee information of multiple employees, historical learning behavior information, historical learning result information, and historical performance levels in the target business knowledge point type; finally, based on the predicted performance level of the target employee in the target business knowledge point type, it is determined whether the target employee poses a business risk. This invention uses the learning behavior and results information of target employees in the target business knowledge point type, along with the corresponding basic employee information, and a trained XGBOOST model to predict the performance level of target employees in the target business knowledge point type. Based on this performance level prediction, it determines whether the target employee has business risks, thereby enabling timely and effective detection of business risks. This solves the problem that existing technologies lack a business risk early warning method based on business knowledge learning, making it impossible to timely and effectively detect business risks caused by insufficient business knowledge learning among employees. Attached Figure Description

[0032] Figure 1 This is a flowchart of a business risk early warning method according to Embodiment 1 of the present invention;

[0033] Figure 2 This is a flowchart of another business risk early warning method according to an embodiment of the present invention;

[0034] Figure 3 This is a schematic diagram of the structure of a business risk early warning device according to Embodiment 2 of the present invention;

[0035] Figure 4 This is a schematic diagram of a business risk early warning device according to Embodiment 3 of the present invention. Detailed Implementation

[0036] To enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0037] It is understood that the specific embodiments and accompanying drawings described herein are merely for explaining the invention and are not intended to limit the invention.

[0038] It is understood that, without conflict, the various embodiments and features in the embodiments of the present invention can be combined with each other.

[0039] It is understood that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, while the parts unrelated to the present invention are not shown in the drawings.

[0040] It is understood that each unit or module involved in the embodiments of the present invention may correspond to only one entity structure, or may be composed of multiple entity structures, or multiple units or modules may be integrated into one entity structure.

[0041] It is understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of this invention may occur in a different order than that marked in the accompanying drawings.

[0042] It is understood that the flowcharts and block diagrams of this invention illustrate the possible architecture, functions, and operations of systems, apparatuses, devices, and methods according to various embodiments of this invention. Each block in the flowchart or block diagram may represent a unit, module, program segment, or code, containing executable instructions for implementing the specified function. Furthermore, each block or combination of blocks in the block diagram and flowchart can be implemented using a hardware-based system to achieve the specified function, or using a combination of hardware and computer instructions.

[0043] It is understood that the units and modules involved in the embodiments of the present invention can be implemented by software or by hardware. For example, the units and modules can be located in a processor.

[0044] Application Overview

[0045] As companies increasingly emphasize continuous learning for their employees, more and more are building internal learning platforms to enhance their professional skills and business acumen. However, most of these platforms are merely used to assess whether employees have completed learning tasks, without effectively analyzing the learning process and outcomes. In business operations, there is a wealth of knowledge in specific vertical fields and subjectively defined concepts. If employees lack access to this type of knowledge, it can lead to significant business risks when conducting marketing or production activities, such as customer churn, data security breaches, and integrity risks.

[0046] Current research on learning data analysis focuses on higher education institutions, primarily aiming to assess the correlation between daily learning behaviors or outcomes and student performance. In contrast to performance prediction in higher education, businesses are primarily profit-driven and therefore place greater emphasis on employee performance and the corresponding prediction of business risks. However, the factors influencing employee performance in businesses are complex and diverse. Research methods used for higher education students are difficult to apply to businesses.

[0047] In summary, existing solutions primarily focus on analyzing and predicting knowledge learning within universities and among students, mainly aiming to improve student performance and optimize teaching methods. There is a lack of a business risk early warning method based on enterprise business knowledge learning, making it difficult to promptly and effectively identify business risks arising from employees' insufficient business knowledge acquisition.

[0048] To address the aforementioned technical problems, this application provides a business risk early warning method, apparatus, and computer-readable storage medium. By utilizing the learning behavior and results information of target employees in target business knowledge point types, along with corresponding basic employee information, a trained XGBOOST model is used to predict the performance level of target employees in those target business knowledge point types. Based on this performance level prediction, it is determined whether the target employee faces business risks. This enables timely and effective detection of business risks, at least solving the problem of the lack of a business risk early warning method based on business knowledge learning in the prior art, which makes it impossible to timely and effectively detect business risks caused by insufficient business knowledge learning among employees.

[0049] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0050] Example 1:

[0051] This embodiment provides a business risk early warning method, such as Figure 1 As shown, the method includes:

[0052] Step S101: Obtain the basic employee information of the target employee, as well as the learning behavior information and learning result information of the target employee in the target business knowledge point type.

[0053] It should be noted that the employee's basic information includes: employee ID, job type, age, length of service, years of service in this position, and education level. The learning behavior information includes: learning start time, number of words in the article, duration of time spent in the article, answering start time, answering duration, number of retries, and average duration of time spent in the article. The learning result information includes: answering score.

[0054] In an optional embodiment, before obtaining the target employee's basic information and the target employee's learning behavior and learning outcome information in the target business knowledge point type, the method further includes:

[0055] Collect uncategorized learning content data and question content data of the target employees in the preset learning platform;

[0056] The unclassified learning content data and the question content data are classified using a trained textCNN model to identify learning content data and question content data under different business knowledge point types.

[0057] Based on the learning content data and question content data under the different business knowledge point types, the learning behavior information and learning result information of the target employee in each business knowledge point type are determined, wherein the different business knowledge point types include the target business knowledge point type.

[0058] Specifically, the system acquires unclassified learning content and question data manually entered by target employees into a pre-developed learning platform. A trained textCNN model is then used to automatically classify the unclassified learning content and question data, automatically generating business knowledge point types for each type of learning content and question in the unclassified learning content and question data. This process identifies the learning content and question data under different business knowledge point types. Based on the learning content and question data under different business knowledge point types, the system determines the learning behavior information and learning results information of target employees when learning and answering questions on the learning content and question data of each business knowledge point type in the learning platform.

[0059] In an optional embodiment, before classifying the unclassified learning content data and the question content data using a trained textCNN model to identify learning content data and question content data under different business knowledge point types, the method further includes:

[0060] Acquire learning content data and question content data that have been tagged with business knowledge point types in the learning platform;

[0061] Based on the learning content data and question content data of the labeled business knowledge point types, a textCNN model is constructed and trained to obtain the trained textCNN model.

[0062] Specifically, the learning content data and question content data, manually entered and labeled with business knowledge point types in the learning platform, are obtained. Taking the basic business knowledge of telecom operators as an example, the business knowledge point types can be divided into mobile network package knowledge, broadband knowledge, marketing strategies, and data package combination knowledge, etc. The learning content data and question content data labeled with business knowledge point types are divided into training set and test set in an 8:2 ratio. An embedding layer is constructed, and word2vec is used to construct word vectors. The text in the training set and test set is converted into word vectors of the same dimension. Based on the text in the training set and test set after being converted into word vectors of the same dimension, a textCNN model is constructed and trained to obtain the trained textCNN model.

[0063] Step S102: Input the employee's basic information, the target employee's learning behavior information and learning result information in the target business knowledge point type into the trained extreme gradient boosting tree XGBOOST model to obtain the predicted performance level of the target employee in the target business knowledge point type. The trained XGBOOST model is trained based on the employee's basic information, historical learning behavior information, historical learning result information and historical performance level of multiple employees in the target business knowledge point type.

[0064] It should be noted that the performance level mentioned is based on the assessment results, which are divided into excellent, average, and poor, with the top 10% being excellent, the middle 80% being average, and the bottom 10% being poor.

[0065] In an optional embodiment, before inputting the employee's basic information, the target employee's learning behavior information and learning result information into the trained extreme gradient boosting tree XGBOOST model to obtain the predicted performance level of the target employee in the target business knowledge point type, the method further includes:

[0066] The employee basic information of the multiple employees, the historical learning behavior information and the historical learning result information in the target business knowledge point type are used as the feature dataset;

[0067] The historical performance level of the multiple employees in the target business knowledge point type is used as the target variable corresponding to the feature dataset;

[0068] Based on the feature dataset and the corresponding target variables, the XGBOOST model is constructed and trained to obtain the trained XGBOOST model.

[0069] Specifically, for each employee among multiple employees, the employee's employee ID is used as the connection point to link the employee's basic information, historical learning behavior information, historical learning result information, and historical performance level in the target business knowledge point type. The basic information, historical learning behavior information, and historical learning result information of all employees in the target business knowledge point type are used as the feature dataset. The historical performance level of all employees in the target business knowledge point type is used as the target variable corresponding to the feature dataset. Data preprocessing is performed on the feature dataset: empty data and employee identifier fields are removed. Non-continuous variables such as job type and education level are labeled and encoded. The preprocessed feature dataset is then divided into training and testing sets in an 8:2 ratio. Based on the training and testing sets, an XGBOOST model is built and trained to obtain the trained XGBOOST model.

[0070] Step S103: Based on the predicted performance level of the target employee in the target business knowledge point type, determine whether the target employee has any business risks.

[0071] Specifically, if the predicted performance level of the target employee in the target business knowledge point type is poor, then it is determined that the target employee poses a business risk.

[0072] In an optional embodiment, if the target employee poses a business risk, the method further includes:

[0073] Relevant measures were taken to provide guidance to the target employees.

[0074] Specifically, for employees with poor performance, relevant measures should be taken in a timely manner to provide guidance, including but not limited to assigning high-performing colleagues to provide one-on-one guidance and increasing the frequency of business knowledge sharing.

[0075] In one specific embodiment, a business risk early warning method and system based on enterprise personnel learning data are provided, the system comprising:

[0076] Online learning platform: A business dictionary is built based on the company's business types, and the dictionary contents are used as tags. An online learning platform is constructed, with professionals responsible for inputting learning materials and questions. Learning content delivery strategies are set based on different tags, and learning tasks are pushed out regularly.

[0077] Data Acquisition Module: This module is responsible for collecting employees' learning behavior data (such as learning duration, learning completion rate, and learning points) and learning outcome data (such as exam scores and accuracy rate) on the learning platform in real time.

[0078] Text recognition module: Based on the first batch of input learning content and answer content (i.e. question content), manual labeling is performed, and textCNN is used to build a model for text classification, so that newly input text content can be automatically classified into specified business tags.

[0079] Business metrics database: Based on the business dictionary, data is collected from the enterprise business platform, and the performance metrics data of the employees to be analyzed are entered into the database.

[0080] Data Analysis Module: This module utilizes big data and machine learning technologies to conduct in-depth correlation analysis on learning behavior data, learning outcome data, and business indicator completion rates, identifying potential correlations between learning data and business performance.

[0081] Risk warning module: Based on data analysis results, it automatically identifies employees with business risks and sends warning information to relevant managers through the system or email.

[0082] The rectification suggestion module provides managers with specific rectification suggestions based on early warning information, combined with employee learning data and business indicator data, to help employees improve their learning methods and enhance business performance.

[0083] like Figure 2 As shown, the specific implementation process of this business risk early warning method includes the following steps:

[0084] Step 1: Develop an online learning platform for the enterprise, and push learning content and quiz content.

[0085] Specifically, develop a learning platform for all employees of the enterprise, enabling functions such as learning subject construction, learning content input, question bank input, push notifications, and management.

[0086] Step 2: Build a text classification model based on textCNN

[0087] Specifically, learning materials and exercises are collected using methods such as manual input and web crawling. The learning content data and exercise data are then cleaned, segmented, corrected, and stored in a database. The stored learning content data and exercise data are manually tagged with their business knowledge point type T. Taking basic business knowledge of telecom operators as an example, business knowledge point types can be categorized as mobile network package knowledge, broadband knowledge, marketing strategies, and data package combination knowledge, etc.

[0088] Specifically, the business knowledge point type T is used as the text category, and a corresponding identifier code is set for T. HTML tags and other special characters are removed from the text, and the data is organized into the format shown in Table 1. In Table 1, "1" and "2" represent the business knowledge point identifier codes.

[0089] Table 1: Text Content and Corresponding Business Knowledge Point Identifier Codes

[0090]

[0091] The data was divided into training and test sets in an 8:2 ratio. An embedding layer was constructed, and word2vec was used to generate word vectors, converting the text into word vectors of uniform dimension. The textCNN natural language model was used to model the training set, constructing convolutional layers with kernels of sizes 3, 4, and 5 sliding across the word embedding matrix, and calculating the convolution results. For the i-th convolutional kernel, its output feature map Z... (i) The formula is as follows, where (W) i ) is the weight matrix of the i-th convolutional kernel, and E represents the input word vector. i ) represents the bias term of the i-th convolution kernel, Conv1D represents a one-dimensional convolution operation, and ReLU is the activation function.

[0092] (Z i =ReLU(ConvlD(W i E)+b (i) ))

[0093] Max pooling is performed, extracting the maximum value of each feature vector to represent that feature. For the i-th feature map, its max pooling result (P) (i) )for:

[0094] (P (i) =MaxPoolinglD(Z (i) ))

[0095] A fully connected layer is constructed to concatenate multiple pooling results; then, classification is performed using the fully connected layer. The formula for the output (O) of the fully connected layer is as follows, where (W... dense ) is the weight matrix of the fully connected layer, and (Concatenate) is the concatenation operation:

[0096] (O = DenseLayer(W) dense Concatenate(P)))

[0097] Based on the output of the fully connected layer, the probability classification of the business type to which the text belongs is obtained, where (W) out (b) is the weight matrix of the output of the fully connected layer. out ) is the bias item of the output of the fully connected layer:

[0098] (Y = softmax(W) out ·O+b out ))

[0099] The learning rate, number of iterations, and other parameters were continuously adjusted, the optimal results were saved, and model training was completed. The test set was then fed into the trained model, and its accuracy, denoted as Accuracyl, was calculated. The Accuracyl accuracy on the test set was 87.6%, a satisfactory result. Accuracy1 is calculated as follows, where TP represents true positives, TN represents true negatives, FP represents false positives, and FN represents false negatives:

[0100]

[0101] Deploying this text classification model enables the learning platform to automatically classify subsequently entered learning content data and question content data, and automatically generate business knowledge point type T for the subsequently entered learning content data and question content data.

[0102] Step 3: Collect employee basic information, learning characteristic information (i.e., learning behavior information), learning performance information (i.e., learning outcome information), and performance information (i.e., performance level).

[0103] Specifically, the data acquisition module in the learning platform is used to collect learning characteristic information Sa of employee U (employee U can be one or more employees) when learning and answering questions on each piece of learning content and question content in the learning platform for business knowledge point type T (business knowledge point type T can be one or more business knowledge point types). (T) And employee basic information Ua.

[0104] Among them, Ua includes the employee's employee ID (uid), job type (gw), age (age), length of service (gl), years of service in this position (gwa), and education level (xl). (T) This includes the employee's learning start time bt for business knowledge point type T, the number of words in the article al, the time spent in the article st, the start time for answering questions qb, the time spent answering questions, the number of retries for answering questions qa, and the average time spent in the article (average time spent per word) aut. Where aut = st / al.

[0105] Specifically, the data collection includes employee U's scores Ga, obtained by answering questions on the learning platform related to business knowledge point type T. (T) Collect employee U's performance evaluation (i.e., performance level) Gs within the company for business knowledge type T. (T) Considering that there may be multiple assessment methods in enterprises, in order to simplify the calculation, the assessment results are divided into three categories: excellent, average, and poor, with the top 10% being excellent, the middle 80% being average, and the bottom 10% being poor.

[0106] Step 4: Merge the collected information to form a feature table, classify the performance information, and use it as the target variable.

[0107] Specifically, using employee U's employee ID as the connection point, connect employee U's Ua and Sa. (T) Ga (T) Gs (T) Summarize them into a single general characteristic table M. Then, assign Ua and Sa to employee U. (T) Ga (T) As a feature dataset, employee U's performance evaluation Gs (T) As the target variable of the feature dataset.

[0108] Step 5: Train and predict using the XGBoost model

[0109] Specifically, data preprocessing is performed to remove empty data and fields such as uid. Label encoding is applied to non-continuous variables such as job type (gw) and education level (xl). The data table M is divided into training and test sets in an 8:2 ratio. An extreme gradient boosting tree (XGBOOST) is used to train and model the training set.

[0110] It should be noted that XGBoost is an additive model based on boosting trees, and its objective function consists of a loss function and a regularization term, where Obj (t) Let (t) represent the total loss value of the current tree, (t) represent the current tree, and n represent the training sample (a user's Ua, Sa). (T) Ga (T) The total number of samples (corresponding to one training sample), y i and These represent the true value of the performance evaluation corresponding to the i-th training sample and the predicted value of the performance evaluation corresponding to the i-th training sample in the current tree, respectively. The loss function represents the difference between the true value corresponding to the i-th training sample and the predicted value corresponding to the i-th training sample in the current tree. Let represent the sum of the complexities of a tree with a total complexity of t. Therefore, the objective function is as follows:

[0111]

[0112] The hyperparameters were continuously adjusted during XGBOOST training to obtain the optimal training results. The model was then validated using test set data, and its accuracy was calculated, yielding Accuracy2. The accuracy rate was 82.12%, a satisfactory result. The XGBOOST model was deployed online, and data was collected at the beginning, middle, and end of each month following the above steps. The XGBOOST model was then used to predict the performance level of employees in the following month.

[0113] Step 6: Based on the forecast results, promptly implement relevant measures to help staff improve performance.

[0114] Specifically, for employees with poor performance, relevant measures should be taken in a timely manner to provide guidance, including but not limited to assigning high-performing colleagues to provide one-on-one guidance and increasing the frequency of business knowledge sharing.

[0115] It is worth mentioning that the business risk early warning method and system based on enterprise personnel learning data provided by this invention can track employees' learning behavior and learning results on the learning platform in real time. Combined with the completion rate of business indicators, it analyzes the correlation between learning data and business performance, thereby accurately determining whether employees pose business risks and issuing timely warnings and providing rectification suggestions. This invention uses modeling based on learning data from the learning platform and employees' job characteristics to prevent business risks caused by insufficient business knowledge learning among enterprise employees. The system collects data such as employees' learning behavior and learning results in real time through the construction of the learning platform, enabling timely and effective detection of business risks.

[0116] The business risk early warning method provided in this embodiment of the invention first obtains the basic employee information of the target employee, as well as the target employee's learning behavior information and learning result information in the target business knowledge point type; then, it inputs the basic employee information, the target employee's learning behavior information and learning result information in the target business knowledge point type into a trained extreme gradient boosting tree (XGBOOST) model to obtain the predicted performance level of the target employee in the target business knowledge point type. The trained XGBOOST model is obtained based on the basic employee information of multiple employees, historical learning behavior information, historical learning result information, and historical performance levels in the target business knowledge point type; finally, it determines whether the target employee poses a business risk based on the predicted performance level of the target employee in the target business knowledge point type. This invention uses the learning behavior and results information of target employees in the target business knowledge point type, along with the corresponding basic employee information, and a trained XGBOOST model to predict the performance level of target employees in the target business knowledge point type. Based on this performance level prediction, it determines whether the target employee has business risks, thereby enabling timely and effective detection of business risks. This solves the problem that existing technologies lack a business risk early warning method based on business knowledge learning, making it impossible to timely and effectively detect business risks caused by insufficient business knowledge learning among employees.

[0117] Example 2:

[0118] like Figure 3 As shown, this embodiment provides a business risk early warning device for executing the above-described business risk early warning method, including:

[0119] The first acquisition module 11 is used to acquire the basic employee information of the target employee, as well as the learning behavior information and learning result information of the target employee in the target business knowledge point type.

[0120] Prediction module 12, connected to the first acquisition module 11, is used to input the employee's basic information, the target employee's learning behavior information and learning result information in the target business knowledge point type into the trained extreme gradient boosting tree XGBOOST model to obtain the predicted value of the target employee's performance level in the target business knowledge point type. The trained XGBOOST model is trained based on the employee's basic information, historical learning behavior information, historical learning result information and historical performance level of multiple employees in the target business knowledge point type.

[0121] The early warning module 13, connected to the prediction module 12, is used to determine whether the target employee has business risks based on the predicted performance level of the target employee in the target business knowledge point type.

[0122] Furthermore, the device also includes:

[0123] The data collection module is used to collect uncategorized learning content data and question content data of the target employees in the preset learning platform;

[0124] The classification module is used to classify the unclassified learning content data and the question content data using a trained textCNN model, so as to identify learning content data and question content data under different business knowledge point types.

[0125] The determination module is used to determine the learning behavior information and learning result information of the target employee in each business knowledge point type based on the learning content data and question content data under the different business knowledge point types, wherein the different business knowledge point types include the target business knowledge point type.

[0126] Furthermore, the device also includes:

[0127] The second acquisition module is used to acquire learning content data and question content data that have been marked with business knowledge point types in the learning platform.

[0128] The first construction and training module is used to construct and train the textCNN model based on the learning content data and question content data of the labeled business knowledge point types, so as to obtain the trained textCNN model.

[0129] Furthermore, the learning behavior information includes: learning start time, article word count, article dwell time, quiz start time, quiz duration, number of quiz retries, and average dwell time per article word count; the learning result information includes: quiz score.

[0130] Furthermore, the device also includes:

[0131] The first module is used to use the basic employee information of the multiple employees, the historical learning behavior information and the historical learning result information in the target business knowledge point type as a feature dataset.

[0132] The second module is used to take the historical performance level of the multiple employees in the target business knowledge point type as the target variable corresponding to the feature dataset.

[0133] The second construction and training module is used to construct and train the XGBOOST model based on the feature dataset and the corresponding target variables, so as to obtain the trained XGBOOST model.

[0134] Furthermore, the performance level is divided into excellent, average, and poor, and the early warning module 13 specifically includes:

[0135] The judgment unit is used to determine that the target employee has business risks if the predicted performance level of the target employee in the target business knowledge point type is poor.

[0136] Furthermore, the early warning module 13 also includes:

[0137] The coaching unit is used to provide coaching and guidance to the target employees by taking relevant measures.

[0138] Example 3:

[0139] refer to Figure 4 This embodiment provides a business risk warning device, including a memory 21 and a processor 22. The memory 21 stores a computer program, and the processor 22 is configured to run the computer program to execute the business risk warning method in Embodiment 1.

[0140] The memory 21 is connected to the processor 22. The memory 21 can be a flash memory, a read-only memory or other memory, and the processor 22 can be a central processing unit or a microcontroller.

[0141] Example 4:

[0142] This embodiment provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the business risk warning method in Embodiment 1 above.

[0143] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules, or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory or other memory technologies, CD-ROM (Compact Disc Read-Only Memory), DVD or other optical disc storage, cartridges, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer.

[0144] In summary, the business risk early warning method, apparatus, and computer-readable storage medium provided in this embodiment of the invention first acquire the basic employee information of the target employee, as well as the target employee's learning behavior information and learning result information in the target business knowledge point type; then, the employee's basic information, the target employee's learning behavior information and learning result information in the target business knowledge point type are input into a trained extreme gradient boosting tree (XGBOOST) model to obtain a predicted performance level of the target employee in the target business knowledge point type. The trained XGBOOST model is obtained based on the basic employee information of multiple employees, historical learning behavior information, historical learning result information, and historical performance levels in the target business knowledge point type; finally, based on the predicted performance level of the target employee in the target business knowledge point type, it is determined whether the target employee poses a business risk. This invention uses the learning behavior and results information of target employees in the target business knowledge point type, along with the corresponding basic employee information, and a trained XGBOOST model to predict the performance level of target employees in the target business knowledge point type. Based on this performance level prediction, it determines whether the target employee has business risks, thereby enabling timely and effective detection of business risks. This solves the problem that existing technologies lack a business risk early warning method based on business knowledge learning, making it impossible to timely and effectively detect business risks caused by insufficient business knowledge learning among employees.

[0145] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A business risk early warning method, characterized in that, The method includes: Obtain basic employee information of the target employee, as well as the target employee's learning behavior and learning outcome information in the target business knowledge point type; The employee's basic information, the target employee's learning behavior information and learning result information in the target business knowledge point type are input into the trained extreme gradient boosting tree XGBOOST model to obtain the predicted performance level of the target employee in the target business knowledge point type. The trained XGBOOST model is trained based on the employee's basic information, historical learning behavior information, historical learning result information and historical performance level of multiple employees in the target business knowledge point type. Based on the predicted performance level of the target employee in the target business knowledge point type, determine whether the target employee poses a business risk.

2. The method according to claim 1, characterized in that, Before obtaining the target employee's basic information and the target employee's learning behavior and learning outcome information in the target business knowledge point type, the method further includes: Collect uncategorized learning content data and question content data of the target employees in the preset learning platform; The unclassified learning content data and the question content data are classified using a trained textCNN model to identify learning content data and question content data under different business knowledge point types. Based on the learning content data and question content data under the different business knowledge point types, the learning behavior information and learning result information of the target employee in each business knowledge point type are determined, wherein the different business knowledge point types include the target business knowledge point type.

3. The method according to claim 2, characterized in that, Before classifying the unclassified learning content data and the question content data using the trained textCNN model to identify learning content data and question content data under different business knowledge point types, the method further includes: Acquire learning content data and question content data that have been tagged with business knowledge point types in the learning platform; Based on the learning content data and question content data of the labeled business knowledge point types, a textCNN model is constructed and trained to obtain the trained textCNN model.

4. The method according to claim 1, characterized in that, The learning behavior information includes: learning start time, article word count, article dwell time, quiz start time, quiz duration, number of quiz retries, and average dwell time per article word count. The learning result information includes: quiz score.

5. The method according to claim 1, characterized in that, Before inputting the employee's basic information, the target employee's learning behavior information and learning result information into the trained extreme gradient boosting tree XGBOOST model to obtain the predicted performance level of the target employee in the target business knowledge point type, the method further includes: The employee basic information of the multiple employees, the historical learning behavior information and the historical learning result information in the target business knowledge point type are used as the feature dataset; The historical performance level of the multiple employees in the target business knowledge point type is used as the target variable corresponding to the feature dataset; Based on the feature dataset and the corresponding target variables, the XGBOOST model is constructed and trained to obtain the trained XGBOOST model.

6. The method according to claim 1, characterized in that, The performance level is categorized as excellent, average, and poor. The determination of whether a target employee poses a business risk based on the predicted performance level of that employee in the target business knowledge point type specifically includes: If the predicted performance level of the target employee for the target business knowledge point type is poor, then the target employee is judged to have business risk.

7. The method according to claim 6, characterized in that, If the target employee poses a business risk, the method further includes: Relevant measures were taken to provide guidance to the target employees.

8. A business risk early warning device, characterized in that, include: The first acquisition module is used to acquire the basic employee information of the target employee, as well as the learning behavior information and learning result information of the target employee in the target business knowledge point type. The prediction module, connected to the first acquisition module, is used to input the employee's basic information, the target employee's learning behavior information and learning result information in the target business knowledge point type into the trained extreme gradient boosting tree XGBOOST model to obtain the predicted value of the target employee's performance level in the target business knowledge point type. The trained XGBOOST model is trained based on the employee's basic information, historical learning behavior information, historical learning result information and historical performance level of multiple employees in the target business knowledge point type. The early warning module, connected to the prediction module, is used to determine whether the target employee has any business risks based on the predicted performance level of the target employee in the target business knowledge point type.

9. A business risk early warning device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to implement the business risk warning method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the business risk warning method as described in any one of claims 1-7.

Images