Intelligent training method, system and equipment for safety production education

Abstract

The invention discloses a safety production education intelligent training method, system and device, relates to the technical field of artificial intelligence, and is used for solving the problems of single safety production education training content, poor interactivity and difficulty in personalized customization in the prior art. Comprising the steps of obtaining safety production related basic data and user portrait data of a target user; constructing a safety production knowledge base based on the safety production related basic data; constructing a safety production education training model based on the safety production knowledge base and the user portrait data; a safety production education training model is adopted to realize intelligent dialogue learning for the target users, and a personalized education training scheme is generated for each target user; and automatically generating an examination question bank according to the personalized education training scheme of each target user. The invention provides a set of complete educational training assessment scheme for safety production, realizes personalized training of safety production of operators, and improves the efficiency and effect of educational training.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an intelligent training method, system and equipment for production safety education. Background Art

[0002] With the rapid development of science and technology, artificial intelligence (AI) technology has gradually penetrated into all walks of life, bringing unprecedented changes to traditional industries. In the field of production safety, traditional education and training methods face problems such as single content, poor interactivity, and difficulty in personalized customization. It is difficult to meet the high requirements of safety awareness and skills of operators in the increasingly complex and changing production environment.

[0003] Traditional safety production education and training mostly adopts centralized teaching or written material learning methods. Although these methods have improved the safety knowledge and skill level of operators to a certain extent, they have many shortcomings. First, the content is single and it is difficult to meet the needs of operators in different positions and with different skill levels; second, the interactivity is poor, and it is impossible to make timely adjustments and optimizations according to the actual situation of the operators; finally, it is difficult to customize and conduct targeted training according to the specific situation of each operator. With the rapid development of AI technology, especially the advantages of large language models in understanding natural language, interacting and generating high-quality content. Therefore, it is particularly important to develop an efficient, flexible and personalized safety production education and training method.

[0004] Therefore, there is an urgent need to provide a more reliable intelligent training solution for production safety education. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent training method, system and equipment for production safety education, which are used to solve the problems in the prior art that the production safety education and training content is single, the interactivity is poor, and it is difficult to customize it.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides an intelligent training method for production safety education, the method comprising:

[0008] Obtaining basic data related to production safety and user portrait data of the target user; the basic data related to production safety at least includes industry standard procedure data and accident case data for various types of work and positions in the production safety industry; the user portrait data at least includes basic user identity data of the target user;

[0009] Building a safety production knowledge base based on the safety production related basic data;

[0010] Constructing a production safety education and training model based on the production safety knowledge base and the user portrait data;

[0011] The production safety education and training model is used to implement intelligent dialogue learning for the target users, and a personalized education and training program is generated for each target user; the personalized education and training program at least includes multimedia education and training content;

[0012] An examination question bank is automatically generated based on the personalized education and training program for each target user.

[0013] Optionally, the building of a safety production knowledge base based on the safety production related basic data specifically includes:

[0014] Structuring the basic data related to production safety to obtain structured data;

[0015] Extracting entities, relationship data, and events from the structured data, and identifying relationships between the extracted entities;

[0016] Based on the relationship, generate personalized graph metadata for constructing a personalized safety production graph corresponding to the user;

[0017] A personalized production safety map corresponding to the user is constructed based on the personalized map metadata to form a production safety knowledge base.

[0018] Optionally, a production safety education and training model is constructed based on the production safety knowledge base and the user portrait data, specifically including:

[0019] Incrementally training the features in the safety production knowledge base by an incremental learning method, and supplementing the sample data in the feature set in the safety production knowledge base by transfer learning to obtain a target industry knowledge feature set;

[0020] Extracting user features of the user portrait data based on a neural network to obtain a user feature set;

[0021] The initial production safety education and training model is trained based on the user feature set corresponding to the user portrait data and the target industry knowledge feature set to obtain a production safety education and training model.

[0022] Optionally, before building the production safety education and training model based on the production safety knowledge base and the user portrait data, the method further includes:

[0023] Real-time acquisition of dynamically updated new data; the new data at least includes dynamically updated user portrait data and newly added safety production related basic data; the sources of the new data at least include user uploaded data, online network updated data and new data from industry documents;

[0024] Based on a preset update mechanism, the production safety knowledge base is updated regularly according to the new data that has passed the content review.

[0025] Optionally, the production safety education and training model is used to implement intelligent dialogue learning for the target users, and a personalized education and training program is generated for each target user, specifically including:

[0026] Obtain user profile information and training course information of the target user; the user profile information of the target user includes at least the basic identity information, historical training records and historical evaluation results of the target user; the training course information includes at least the course introduction, difficulty level, training time and knowledge points;

[0027] Automatically generate question guidance information based on the user portrait information of the target user; the question guidance information at least includes ToolboxTalk mode prompt words;

[0028] Collecting questions or needs raised by the target user according to the question guiding information, and determining personalized information of the target user based on the questions or needs; the personalized information at least includes the course preference of the target user;

[0029] Extracting a first feature vector corresponding to the user portrait information of the target user and a second feature vector corresponding to the personalized information;

[0030] Based on the first feature vector and the second feature vector, the dialogue strategy and recommended content are adjusted based on the safety production education and training model dynamically combined with collaborative filtering or content filtering methods to generate a personalized education and training plan.

[0031] Optionally, the adopting of the production safety education and training model to implement intelligent dialogue learning for the target users and generating a personalized education and training program for each target user also includes:

[0032] Performing sentiment analysis on the training course information to identify sentiment tags corresponding to the training course information;

[0033] Using a support vector machine combined with the emotional tags to perform emotional classification on the training course information, and constructing an emotional database;

[0034] Generate personalized multimedia education and training content for the target user, and dynamically adjust the multimedia education and training content in combination with the emotion database; the multimedia education and training content includes at least animation, video, text, audio and picture; the dynamic adjustment includes at least adjusting the tone, speech speed, sound effects, background music, animation character expression or action.

[0035] Optionally, after generating personalized multimedia education and training content for the target user, the method further includes:

[0036] Acquire interactive behavior information between the target user and the multimedia education and training content; the interactive behavior information at least includes click information, comment information and viewing time of the target user on the multimedia education and training content;

[0037] Based on the interactive behavior information and combined with reinforcement learning, the multimedia education and training content is dynamically adjusted; the multimedia education and training content supports exporting and sharing in multiple media formats.

[0038] Optionally, the step of automatically generating a test question bank according to the personalized education and training program for each target user specifically includes:

[0039] Automatically generate assessment questions based on the production safety knowledge base, the training course information and pre-stored question templates;

[0040] Based on the target user's learning progress for the personalized education and training program and the interactive behavior information, automatically adjust the difficulty of the assessment questions and generate an assessment question bank for the target user;

[0041] After generating the test question bank for the target user, the method further includes:

[0042] Obtaining the target user's assessment score for the assessment question bank, performing intelligent assessment for the target user, obtaining the assessment result, and collecting feedback data from the target user;

[0043] Based on the evaluation results and the feedback data, the production safety education and training model is updated.

[0044] Compared with the prior art, the present invention provides an intelligent training method for safe production education, which obtains basic data related to safe production and user portrait data of target users; builds a safe production knowledge base based on the basic data related to safe production; builds a safe production education and training model based on the safe production knowledge base and user portrait data; uses the safe production education and training model to realize intelligent dialogue learning for target users, and generates a personalized education and training plan for each target user; and automatically generates an assessment question bank based on the personalized education and training plan for each target user. The present invention provides a complete education, training and assessment plan for safe production, realizes personalized training for safe production of operators, improves the efficiency and effect of education and training, and has broad application prospects and promotion value.

[0045] In a second aspect, the present invention provides an intelligent training system for production safety education, the device comprising:

[0046] Safety production knowledge base, personalized education and training program generation module, and assessment question bank generation module;

[0047] The safety production knowledge base is constructed from basic data related to safety production;

[0048] The personalized education and training program generation module builds a safety production education and training model based on the safety production knowledge base and user portrait data; and uses the safety production education and training model to implement intelligent dialogue learning for the target user, and generates a personalized education and training program for each target user; the personalized education and training program at least includes multimedia education and training content; the safety production related basic data at least includes industry standard procedure data and accident case data for various types of work and various positions in the safety production industry; the user portrait data at least includes the basic user identity data of the target user;

[0049] The assessment question bank generation module is used to automatically generate an assessment question bank according to the personalized education and training program of each target user.

[0050] In a third aspect, the present invention provides an intelligent training device for production safety education, the device comprising:

[0051] A memory, a processor, and a communication interface coupled to the processor; the memory stores a computer program that can be executed by the processor; when the processor runs the computer program, it executes the intelligent training method for safe production education as provided in the first aspect.

[0052] The technical effects achieved by the system-type solutions provided in the second aspect and the equipment-type solutions provided in the third aspect are the same as those of the method-type solutions provided in the first aspect, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0054] Figure 1 A flowchart of an intelligent training method for production safety education provided by the present invention;

[0055] Figure 2 A schematic diagram of the structure of an intelligent training system for production safety education provided by the present invention;

[0056] Figure 3 A schematic diagram of an intelligent training device for production safety education provided by the present invention. DETAILED DESCRIPTION

[0057] In order to clearly describe the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, words such as "first" and "second" are used to distinguish the same items or similar items with basically the same functions and effects. For example, the first threshold and the second threshold are only used to distinguish different thresholds, and their order is not limited. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit them to be different.

[0058] It should be noted that, in the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0059] In the present invention, "at least one" means one or more, and "plurality" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the associated objects before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, the combination of a and b, the combination of a and c, the combination of b and c, or the combination of a, b and c, where a, b, c can be single or multiple.

[0060] In the field of safe production, traditional education and training methods face problems such as single content, poor interactivity, and difficulty in personalization. They are unable to meet the high requirements of safety awareness and skills of operators in the increasingly complex and changing production environment.

[0061] Therefore, the present invention proposes a method to build a professional safety production knowledge base for specific industries. By establishing a personalized intelligent body, various types of work and positions in safety production enterprises can improve the safety awareness and skills of operators through intelligent dialogue and personalized education and training. The safety production education and training method based on AI technology aims to improve the efficiency and effectiveness of education and training through advanced technical means, and escort the safety production of enterprises.

[0062] Next, the solution provided by the embodiments of this specification is described in conjunction with the accompanying drawings:

[0063] like Figure 1 As shown, the process may include the following steps:

[0064] Step 110: Obtain basic data related to production safety and user profile data of target users;

[0065] The basic data related to production safety at least include industry standard and procedural data and accident case data for various types of work and positions in the production safety industry; specifically, the basic data related to production safety refers to the comprehensive collection of laws, regulations, standards, specifications, operating procedures, accident cases and other data related to production safety for various types of work and positions in a specific industry for production safety, and the data are organized in a structured manner. These data will serve as the basis of the production safety knowledge base and provide strong support for subsequent training.

[0066] The user portrait data at least includes the basic user identity data of the target user. Specifically, the user portrait data may include the user's basic user identity data (such as name, age, occupation, etc.), user preferences, living habits, user behavior, and other labeled data.

[0067] Structural organization may include operations such as classifying and labeling data, and extracting key information.

[0068] This step mainly involves collecting various data related to production safety, including but not limited to laws and regulations, standards and specifications, operating procedures, and accident cases. After collection, these data need to be structured for subsequent processing.

[0069] Step 120: Building a production safety knowledge base based on the production safety related basic data.

[0070] The safety production knowledge base may be composed of a safety production graph. The safety production graph may be a knowledge graph composed of basic data related to safety production.

[0071] Step 130: Construct a production safety education and training model based on the production safety knowledge base and the user portrait data.

[0072] The production safety education and training model can be a deep learning model, and the production safety education model can be used to provide personalized education and training programs for workers in the production safety field.

[0073] When training the production safety education and training model, the training can be based on the data in the production safety knowledge base and the user portrait data. Of course, in the actual implementation process, the training samples of the production safety education and training model include not only the data in the production safety knowledge base and the user portrait data, but also historical training data of the production safety industry, historical feedback data of users, and so on.

[0074] Step 140: Use the production safety education and training model to implement intelligent dialogue learning for the target users, and generate a personalized education and training plan for each target user; the personalized education and training plan at least includes multimedia education and training content.

[0075] Intelligent dialogue is mainly used to collect users' learning needs and questions raised by users, so as to generate personalized education and training plans for users.

[0076] The generation of personalized education and training programs includes intelligent recommendation algorithms, multimedia content generation, and ToolboxTalk mode prompt word design. Among them, ToolboxTalk mode prompt words: use the ToolboxTalk management mode in safety management to achieve interactive dialogue in the intelligent agent under specific safety education and training topics and safety meeting topics, and design structured safety education and training prompt words. By setting specific prompt words and dialogue processes, guide operators to actively participate in discussions and exchanges, and share experiences and lessons. Realize more vivid and effective education and training and safety meetings (including safety morning meetings, safety briefing meetings, etc.) assisted by intelligent agents.

[0077] Step 150: Automatically generate an assessment question bank based on the personalized education and training program for each target user.

[0078] After the target user has learned the personalized education and training program, he or she needs to take an assessment. The present invention provides a complete education and training assessment program for safe production.

[0079] Figure 1The method in the invention obtains basic data related to safe production and user portrait data of target users; constructs a safe production knowledge base based on the basic data related to safe production; constructs a safe production education and training model based on the safe production knowledge base and user portrait data; uses the safe production education and training model to realize intelligent dialogue learning for target users, and generates a personalized education and training plan for each target user; and automatically generates an assessment question bank based on the personalized education and training plan for each target user. The present invention provides a complete education, training and assessment plan for safe production, realizes personalized training for safe production of operators, improves the efficiency and effect of education and training, and has broad application prospects and promotion value.

[0080] based on Figure 1 The method, the embodiments of this specification also provide some specific implementation methods of the method, which are described below.

[0081] Step 120 may specifically include:

[0082] Structuring the basic data related to production safety to obtain structured data;

[0083] Extracting entities, relationship data, and events from the structured data, and identifying relationships between the extracted entities;

[0084] Based on the relationship, generate personalized graph metadata for constructing a personalized safety production graph corresponding to the user;

[0085] A personalized production safety map corresponding to the user is constructed based on the personalized map metadata to form a production safety knowledge base.

[0086] Among them, entities, relationship data and events are extracted from the structured data, and the relationships between the extracted entities are identified. In the specific implementation, the key nouns in the structured data can be used as entities, the associated words or phrases in the structured data can be used as relationships, and the event descriptions in the structured data can be used as events in combination with the actual application scenarios.

[0087] For example, an entity can refer to various nouns related to safe production, such as equipment names, operation names, etc.; a relationship can refer to the association between entities, such as "equipment A should be operated using operating procedure B"; an event may refer to an accident or operation related to safe production, etc.

[0088] In this step, AI technology, especially natural language processing technology, is used to convert the collected data into a machine-readable knowledge base, including entities, relationships, events and other elements.

[0089] In step 130, a production safety education and training model is constructed based on the production safety knowledge base and the user portrait data, which may specifically include:

[0090] Incrementally training the features in the safety production knowledge base by an incremental learning method, and supplementing the sample data in the feature set in the safety production knowledge base by transfer learning to obtain a target industry knowledge feature set;

[0091] Extracting user features of the user portrait data based on a neural network to obtain a user feature set;

[0092] The initial production safety education and training model is trained based on the user feature set corresponding to the user portrait data and the target industry knowledge feature set to obtain a production safety education and training model.

[0093] In the above steps, incremental learning refers to constantly learning new knowledge from new samples and preserving most of the previously learned knowledge. Knowledge is updated gradually, and previous knowledge can be corrected and strengthened so that the updated knowledge can adapt to the newly arrived data without having to relearn all the data. Many machine learning algorithms can apply incremental learning, such as decision trees, rule learning, neural networks (RBF networks, Learn++, FuzzyARTMAP, TopoART, IGNG) and incremental SVM. The learn++ algorithm is an algorithm suitable for supervised learning, integrated, incremental learning, and capable of learning new classes.

[0094] In addition, in this solution, the number of labeled training samples in the safety production industry is limited. Moreover, the data distribution in the field of safety production is related to time, location or other dynamic factors. As the dynamic factors change, the data distribution will change. The previously collected data is outdated and needs to be re-collected and the model rebuilt. Therefore, the present invention uses knowledge transfer to supplement the training sample data.

[0095] In the specific implementation process, a neural network can be used to extract features from the user portrait data. In practical applications, the user portrait data or the data in the safety production knowledge base may contain image data. The data can be convolved and pooled first to extract the features of the characters. Then, LSTM extracts the characters in the image based on the features of the characters and the preset character feature set, and finally obtains the recognition and classification results of the characters to complete the target detection of the characters. By constructing a large amount of sample data through transfer learning to supplement the character feature set with a large amount of sample data, it can not only make up for the large number of samples required for character recognition, but also improve the recognition effect.

[0096] It should be noted that the methods and specific algorithms of transfer learning can be selected according to the actual application needs. The transfer learning methods can include sample transfer, feature transfer, parameter (model) transfer, relationship transfer, etc. The specific algorithm of transfer learning can choose the existing algorithm. Therefore, no specific limitation is made to transfer learning here.

[0097] When training the safety production education and training model, a large number of data samples (data in the safety production knowledge base, user portrait data, historical training data of the safety production industry, historical feedback data of users, etc.) processed by incremental learning and transfer learning can be used for training. During training, a large number of data samples processed by incremental learning and transfer learning are first preprocessed, for example: the entities, relationships and events extracted from the data in the safety production knowledge base are converted into a format that the model can understand, such as vector representation, graph structure, etc., and then the preprocessed data samples are input into the neural network. Each neuron first inputs the value weighted accumulation and then inputs the activation function as the output value of the neuron. The result is obtained by forward propagation, and the result is input into the error function. The error is compared with the expected value to obtain the error. The gradient vector is determined based on the error through back propagation (reverse derivation, error function and each activation function in the neural network are required, and the ultimate goal is to minimize the error). Finally, each weight is adjusted through the gradient vector to make the error tend to 0 or convergence trend adjustment until the set number of times or the average value of the loss error is stable and the training is completed.

[0098] In addition, when training the safety production education and training model, you can choose a large model suitable for processing natural language to carry out safety production knowledge training, such as the GPT series, BERT, etc. These models have been pre-trained on large-scale text data and have strong language understanding and generation capabilities. Fine-tune the model according to the needs of specific industries to ensure that the model can accurately understand industry terms and specific contexts.

[0099] Optionally, before building the production safety education and training model based on the production safety knowledge base and the user portrait data, the following steps may also be included:

[0100] Real-time acquisition of dynamically updated new data; the new data at least includes dynamically updated user portrait data and newly added safety production related basic data; the sources of the new data at least include user uploaded data, online network updated data and new data from industry documents;

[0101] Based on a preset update mechanism, the production safety knowledge base is updated regularly according to the new data that has passed the content review.

[0102] When updating the safety production knowledge base, you can update it based on the following update methods:

[0103] 1) Set the update frequency:

[0104] Regular evaluation: First, you need to determine the frequency of knowledge base updates, which can be monthly, quarterly, or annual, depending on the speed of industry changes, the frequency of regulatory updates, and changes in the production environment.

[0105] Flexibility: In some cases, such as when a major security incident occurs or when regulations change, it may be necessary to temporarily increase the update frequency.

[0106] 2) Content review and screening:

[0107] Data sources: Identify reliable data sources, such as industry reports, government announcements, academic research, professional journals, etc.

[0108] Content screening: Screening newly collected data to ensure its accuracy, relevance, and usefulness, and to avoid the introduction of outdated or erroneous information.

[0109] 3) Update process:

[0110] Assign responsibilities: Appoint a dedicated knowledge base administrator or team to be responsible for updating and maintaining content.

[0111] Review process: Establish a rigorous review process to ensure that all new or modified content undergoes at least one round of peer review or expert review.

[0112] Version control: Implement version control to record the content, time and person in charge of each update for easy tracking and backtracking.

[0113] 4) User feedback mechanism:

[0114] Feedback channels: Provide user feedback channels, such as online forms, email, or forums, to encourage users to report bugs, make suggestions for improvements, or share newly discovered safety information.

[0115] Response mechanism: Establish a rapid response mechanism to handle user feedback in a timely manner and adjust the knowledge base content based on the feedback.

[0116] 5) Training and notification:

[0117] Internal training: Regularly train knowledge base administrators to improve their information screening, content editing and review capabilities.

[0118] User notification: After a major update to the knowledge base, users will be notified via email, announcements, or system messages to ensure they are aware of the latest production safety information.

[0119] 6) Continuous evaluation and improvement:

[0120] Effect evaluation: Regularly evaluate the effect of knowledge base updates, including user satisfaction, content usage rate, error rate and other indicators.

[0121] Continuous optimization: Based on the evaluation results, continuously adjust and update strategies and optimize processes to improve the quality and efficiency of the knowledge base.

[0122] In summary, the regular update mechanism is a process involving multiple links and continuous optimization, which requires clear time planning, strict content review, effective user feedback, timely training notifications, and continuous evaluation and improvement. Through such a mechanism, it can ensure that the knowledge base is always kept up to date and provide strong support for safe production.

[0123] In the above update plan, a regular update mechanism is established to ensure the timeliness and accuracy of the knowledge base content. By continuously collecting and integrating the latest production safety data, updating and improving the knowledge base, the training content is always consistent with the actual production environment.

[0124] Preferably, the production safety education and training model is used to implement intelligent dialogue learning for the target users, and a personalized education and training program is generated for each target user, which may specifically include:

[0125] Obtain user profile information and training course information of the target user; the user profile information of the target user includes at least the basic identity information, historical training records and historical evaluation results of the target user; the training course information includes at least the course introduction, difficulty level, training time and knowledge points;

[0126] Automatically generate question guidance information based on the user portrait information of the target user; the question guidance information at least includes ToolboxTalk mode prompt words;

[0127] Collecting questions or needs raised by the target user according to the question guiding information, and determining personalized information of the target user based on the questions or needs; the personalized information at least includes the course preference of the target user;

[0128] Extracting a first feature vector corresponding to the user portrait information of the target user and a second feature vector corresponding to the personalized information;

[0129] Based on the first feature vector and the second feature vector, the dialogue strategy and recommended content are adjusted based on the safety production education and training model dynamically combined with collaborative filtering or content filtering methods to generate a personalized education and training plan.

[0130] Performing sentiment analysis on the training course information to identify sentiment tags corresponding to the training course information;

[0131] Using a support vector machine combined with the emotional tags to perform emotional classification on the training course information, and constructing an emotional database;

[0132] Generate personalized multimedia education and training content for the target user, and dynamically adjust the multimedia education and training content in combination with the emotion database; the multimedia education and training content includes at least animation, video, text, audio and picture; the dynamic adjustment includes at least adjusting the tone, speech speed, sound effects, background music, animation character expression or action.

[0133] Acquire interactive behavior information between the target user and the multimedia education and training content; the interactive behavior information at least includes click information, comment information and viewing time of the target user on the multimedia education and training content;

[0134] Based on the interactive behavior information and combined with reinforcement learning, the multimedia education and training content is dynamically adjusted; the multimedia education and training content supports exporting and sharing in multiple media formats.

[0135] Among them, natural language processing (NLP) technology: using advanced NLP technology, including intent recognition, entity extraction, semantic understanding, etc., to achieve accurate analysis of the operator's conversation content. Through natural language processing technology, the system can accurately understand the operator's questions and needs, thereby providing more accurate and personalized training services.

[0136] Dialogue interaction design: Design a reasonable dialogue process to guide operators to gradually expose their safety awareness and skill levels. At the same time, use user profiling technology to dynamically adjust dialogue strategies and recommended content based on the background information of operators (such as type of work, position, length of service, etc.) and the content of the dialogue. Implement multi-round dialogue functions to allow the system to gradually gain a deeper understanding of users' needs based on their feedback.

[0137] Safety awareness and skills assessment: Use AI language big models to analyze the conversation content of operators and assess their safety awareness and skill levels. Through deep learning and natural language processing technology, the system can accurately identify the knowledge blind spots or misunderstandings of operators and give corresponding prompts and suggestions.

[0138] Automatically generate assessment questions based on the production safety knowledge base, the training course information and pre-stored question templates;

[0139] Based on the target user's learning progress for the personalized education and training program and the interactive behavior information, automatically adjust the difficulty of the assessment questions and generate an assessment question bank for the target user;

[0140] After generating the test question bank for the target user, the method further includes:

[0141] Obtaining the target user's assessment score for the assessment question bank, performing intelligent assessment for the target user, obtaining the assessment result, and collecting feedback data from the target user;

[0142] Based on the evaluation results and the feedback data, the production safety education and training model is updated.

[0143] Next, the above personalized education and training program generation plan can be described in detail:

[0144] 1) Data collection and preprocessing:

[0145] User profile information of target users: collect basic information of operators (such as age, gender, position, etc.), past training records, evaluation results, etc.

[0146] Training course information: Collect all available training course information, including course introduction, difficulty level, time required, knowledge points involved, etc.

[0147] Learning preferences: Understand the learning style of the operator (e.g., visual, auditory, hands-on), preferred topics, etc.

[0148] Preprocessing: Clean data, handle missing values, outliers, etc., to ensure data quality.

[0149] 2) Feature extraction:

[0150] User characteristics: Construct a feature vector that reflects the user's background, capabilities, and preferences.

[0151] Course characteristics: Construct a feature vector that reflects the course content, difficulty, and relevance of knowledge points.

[0152] Knowledge: Graph: Build a relationship graph between entities to understand the dependencies between different entities.

[0153] 3) User portrait construction:

[0154] User profiles: Combine user data and learning preferences to create a detailed profile for each user, including their interests, skill level, learning style, etc.

[0155] Dynamic updates: Update user profiles regularly or in real time to reflect the user's learning progress and changing interests.

[0156] 4) Model selection and training:

[0157] Collaborative filtering: Recommend similar courses based on user historical behavior and ratings.

[0158] Content filtering: Recommendations are made based on the match between course content and user profile.

[0159] Deep learning models: Use neural networks (such as convolutional neural networks (CNN), recurrent neural networks (RNN), or Transformer) to capture complex user-course interaction patterns.

[0160] Reinforcement learning: enables it to optimize the recommendation strategy through interaction with the environment.

[0161] 5) Comprehensive recommendation strategy:

[0162] Personalized ranking: Calculate a personalized score for each course based on user profile and course characteristics.

[0163] Variety and novelty: Make sure your recommendation list includes courses of different types and difficulty levels to increase variety and novelty.

[0164] Challenge and fun: Consider the user's current knowledge level and learning ability, balance the difficulty of the recommended courses, and consider the user's interests to increase fun.

[0165] 6) Experiment and evaluation:

[0166] A / B testing: Experiment with different recommendation strategies and compare their effectiveness.

[0167] Feedback loop: Collect user feedback on recommended courses to adjust and optimize the recommendation algorithm.

[0168] Evaluation indicators: Use accuracy, recall, F1 score, average precision and other indicators to measure the recommendation effect.

[0169] 7) Deployment and maintenance:

[0170] Online deployment: Integrate the recommendation system into existing education and training platforms.

[0171] Continuous monitoring: Monitor the performance of the recommendation system and adjust model parameters regularly.

[0172] User feedback: Collect user feedback and continuously improve the user experience.

[0173] The multimedia content generated can be high-quality multimedia education and training content. For example, it can include animated videos with automatic dubbing, simulation exercises based on real scenes, etc. By introducing emotional computing technology, the content is made more vivid and interesting, which improves the user's learning interest and participation. At the same time, it supports the export and sharing of multiple media formats, which is convenient for users to learn anytime and anywhere.

[0174] The introduction of emotional computing technology into multimedia content generation aims to enhance the emotional expression of the content, making it closer to real human communication, thereby increasing user participation and learning interest. The following is the specific implementation process:

[0175] 1) Sentiment Analysis:

[0176] First, conduct sentiment analysis on existing high-quality education and training content to identify emotional labels such as happiness, sadness, surprise, anger, etc.

[0177] Use machine learning algorithms (such as SVM, neural networks, etc.) to classify the sentiment of text, audio and video content and establish a sentiment database.

[0178] 2) Emotion model construction:

[0179] Based on the results of sentiment analysis, a sentiment model is constructed. This model can understand and simulate human emotional responses and generate appropriate emotional expressions based on the context of the content.

[0180] The model can include a sentiment rule base, a sentiment lexicon, and a sentiment reasoning mechanism to ensure that the generated content is emotionally coherent and logical.

[0181] 3) Content generation and emotional integration:

[0182] When generating multimedia content (e.g., animated videos, simulation drills, etc.), emotion models are combined with content generation engines.

[0183] Based on the theme and plot of the content, the emotion model provides appropriate emotional guidance, such as adjusting the expressions and movements of animated characters, or changing the tone and speed of dubbing.

[0184] 4) Real-time emotional adjustment:

[0185] When users interact with multimedia content, the emotional expression of the content is adjusted in real time through user feedback and behavior data (such as click-through rate, viewing time, comments, etc.).

[0186] Use methods such as reinforcement learning to continuously optimize the sentiment model to make it more in line with users' emotional needs and preferences.

[0187] 5) Multimodal emotional expression:

[0188] Combine multiple media formats such as text, audio, and video to achieve multimodal emotional expression.

[0189] For example, in an animated video, emotions are conveyed not only through the expressions and movements of the characters, but also through background music, sound effects, and text descriptions to enhance the expression of emotions.

[0190] 6) User testing and feedback:

[0191] Conduct user testing on the generated multimedia content and collect user feedback on emotional expression.

[0192] The sentiment model is fine-tuned based on user feedback to ensure that the generated content is more emotionally aligned with user expectations and needs.

[0193] 7) Export and share:

[0194] Supports export in multiple media formats, such as MP4, AVI, MOV and other video formats, and MP3, WAV and other audio formats.

[0195] Provide convenient sharing functions, allowing users to easily share multimedia content to social media, learning platforms or other channels.

[0196] Through the above steps, affective computing technology can be effectively introduced into multimedia content generation, thereby generating more vivid, interesting, and emotional high-quality educational and training content.

[0197] In the above method, a personalized education and training program is generated for the operator based on the evaluation results and learning preferences of the operator. The algorithm comprehensively considers factors such as the relevance and difficulty of knowledge points and the user's learning preferences to ensure that the training program not only meets the actual needs of the user, but also has a certain degree of challenge and interest.

[0198] In addition, automatic assessment question bank generation and data feedback may include the following implementation steps:

[0199] Question bank generation algorithm: Develop an intelligent question bank generation algorithm to automatically generate test questions based on the knowledge points and question templates in the knowledge base. The algorithm can ensure the diversity, coverage and reasonable distribution of difficulty of the questions, and meet the test needs of operators at different levels and needs.

[0200] Difficulty adjustment mechanism: Automatically adjust the difficulty of questions according to the learning progress and mastery level of the operators to ensure the pertinence and effectiveness of the assessment.

[0201] Instant feedback and analysis: After the assessment is completed, the system provides instant feedback and detailed analysis to help operators consolidate knowledge points. AI technology is used to intelligently evaluate the assessment results of operators, and quickly give scores and analysis. Based on the assessment results, personalized learning suggestions and improvement directions are provided to operators.

[0202] Data feedback loop: Establish a data feedback mechanism to collect user feedback data (such as satisfaction, learning outcomes, etc.) during use to continuously optimize system performance. Regularly analyze user behavior data to identify potential problems and needs and guide the iterative upgrade of system functions.

[0203] The solution of the present invention, the professional safety production knowledge base is realized through data collection and collation, knowledge base construction and knowledge base update mechanism, and also includes dialogue interaction design, safety awareness and skill assessment and personalized feedback. Automatic assessment question bank generation includes question bank generation algorithm, difficulty adjustment mechanism and instant feedback and analysis. Specific technical effects may include:

[0204] 1) Personalized customization: Provide personalized education and training programs based on the actual needs and levels of operators to improve training results.

[0205] 2) Efficient interaction: Through the intelligent dialogue system, real-time interaction with operators can be achieved to enhance learning participation and interest.

[0206] 3) Dynamic update: The knowledge base and examination question bank can be updated regularly to ensure the timeliness and accuracy of the training content.

[0207] 4) Intelligent evaluation and feedback: Use AI technology to achieve intelligent evaluation and feedback to improve the accuracy and timeliness of evaluation.

[0208] Based on the same idea, the present invention also provides an intelligent training system for production safety education, such as Figure 2 As shown, a safety production education intelligent training system 200 may include:

[0209] A safety production knowledge base 210, a personalized education and training program generation module 220, and an assessment question bank generation module 230;

[0210] The safety production knowledge base 210 is constructed from basic data related to safety production;

[0211] The personalized education and training program generation module 220 builds a safety production education and training model based on the safety production knowledge base and user portrait data; and uses the safety production education and training model to implement intelligent dialogue learning for the target user, and generates a personalized education and training program for each target user; the personalized education and training program at least includes multimedia education and training content; the safety production related basic data at least includes industry standard procedure data and accident case data for various types of work and various positions in the safety production industry; the user portrait data at least includes the basic user identity data of the target user;

[0212] The assessment question bank generation module 230 is used to automatically generate an assessment question bank according to the personalized education and training program for each target user.

[0213] Through the collaborative work of the safety production knowledge base, the personalized education and training program generation module, and the assessment question bank generation module, efficient and personalized training of operators' safety awareness and skills can be achieved. By building a professional safety production knowledge base for specific industries and establishing personalized intelligent entities, various types of work and positions in safety production enterprises can improve the safety awareness and skills of operators through intelligent dialogue and personalized education and training.

[0214] Based on the same idea, the embodiments of this specification also provide an intelligent training device for safe production education. Figure 3 As shown, the device includes:

[0215] A memory, a processor, and a communication interface coupled to the processor; the memory stores a computer program that can be executed by the processor; when the processor runs the computer program, the aforementioned intelligent training method for safe production education is executed.

[0216] like Figure 3 As shown, the processor may be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present invention. The communication interface may be one or more. The communication interface may use any transceiver-like device for communicating with other devices or communication networks.

[0217] like Figure 3 As shown, the terminal device may further include a communication line. The communication line may include a path to transmit information between the components.

[0218] Optional, such as Figure 3 As shown, the terminal device may further include a memory. The memory stores a computer program that can be executed by the processor; when the processor executes the computer program, the method provided by the embodiment of the present invention is implemented.

[0219] like Figure 3As shown, the memory may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may exist independently and be connected to the processor through a communication line. The memory may also be integrated with the processor.

[0220] Optionally, the computer-executable instructions in the embodiment of the present invention may also be referred to as application program codes, which is not specifically limited in the embodiment of the present invention.

[0221] In a specific implementation, as an example, Figure 3 As shown, the processor may include one or more CPUs, such as Figure 3 CPU0 and CPU1 in.

[0222] In a specific implementation, as an example, Figure 3 As shown, the terminal device may include multiple processors, such as Figure 3 Each of these processors can be a single-core processor or a multi-core processor.

[0223] Based on the same idea, the embodiments of this specification also provide a computer storage medium corresponding to the above embodiments, and the computer storage medium stores instructions, and when the instructions are executed, the method in the above embodiments is implemented.

[0224] The above mainly introduces the solution provided by the embodiment of the present invention from the perspective of the interaction between the various modules. It is understandable that, in order to realize the above functions, each module includes a hardware structure and / or software unit corresponding to the execution of each function. It should be easily appreciated by those skilled in the art that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present invention can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0225] The embodiment of the present invention can divide the functional modules according to the above method example. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the embodiment of the present invention is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0226] The processor in this specification may also have the function of a memory. The memory is used to store computer-executable instructions for executing the solution of the present invention, and the execution is controlled by the processor. The processor is used to execute the computer-executable instructions stored in the memory, thereby implementing the method provided by the embodiment of the present invention.

[0227] The memory may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may exist independently and be connected to the processor through a communication line. The memory may also be integrated with the processor.

[0228] Optionally, the computer-executable instructions in the embodiment of the present invention may also be referred to as application program codes, which is not specifically limited in the embodiment of the present invention.

[0229] The method disclosed in the above embodiment of the present invention can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor may be a general-purpose processor, a digital signal processor (digital signal processing, DSP), an ASIC, a field-programmable gate array (field-programmable gate array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in the embodiment of the present invention can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0230] Although the present invention is described herein in conjunction with various embodiments, in the process of implementing the claimed invention, those skilled in the art may understand and implement other variations of the disclosed embodiments by viewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "one" or "an" does not exclude multiple situations. A single processor or other unit may implement several functions listed in the claims. Certain measures are recorded in different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0231] Although the present invention has been described in conjunction with specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present invention. Accordingly, this specification and the accompanying drawings are merely exemplary illustrations of the present invention as defined by the appended claims and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present invention. Obviously, those skilled in the art may make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, the present invention is intended to include such modifications and variations if they fall within the scope of the claims of the present invention and their equivalents.

Claims

1. A method for intelligent training of production safety education, characterized in that: Methods include: Obtaining basic data related to safe production and user portrait data of target users; the basic data related to safe production at least includes industry standard procedure data and accident case data for various types of work and positions in the safe production industry; The user portrait data at least includes basic user identity data of the target user; Building a safety production knowledge base based on the safety production related basic data; Constructing a production safety education and training model based on the production safety knowledge base and the user portrait data; The production safety education and training model is used to implement intelligent dialogue learning for the target users, and a personalized education and training program is generated for each target user; the personalized education and training program at least includes multimedia education and training content; An examination question bank is automatically generated based on the personalized education and training program for each target user.

2. The intelligent training method for production safety education according to claim 1, characterized in that: The construction of a safety production knowledge base based on the safety production related basic data specifically includes: Structuring the basic data related to production safety to obtain structured data; Extracting entities, relationship data, and events from the structured data, and identifying relationships between the extracted entities; Based on the relationship, generate personalized graph metadata for constructing a personalized safety production graph corresponding to the user; A personalized production safety map corresponding to the user is constructed based on the personalized map metadata to form a production safety knowledge base.

3. The intelligent training method for production safety education according to claim 1, characterized in that: Building a production safety education and training model based on the production safety knowledge base and the user portrait data specifically includes: Incrementally training the features in the safety production knowledge base by an incremental learning method, and supplementing the sample data in the feature set in the safety production knowledge base by transfer learning to obtain a target industry knowledge feature set; Extracting user features of the user portrait data based on a neural network to obtain a user feature set; The initial production safety education and training model is trained based on the user feature set corresponding to the user portrait data and the target industry knowledge feature set to obtain a production safety education and training model.

4. The intelligent training method for production safety education according to claim 1, characterized in that: Before building the production safety education and training model based on the production safety knowledge base and the user portrait data, the method further includes: Real-time acquisition of dynamically updated new data; the new data at least includes dynamically updated user portrait data and newly added safety production related basic data; the sources of the new data at least include user uploaded data, online network updated data and new data from industry documents; Based on a preset update mechanism, the production safety knowledge base is updated regularly according to the new data that has passed the content review.

5. The intelligent training method for production safety education according to claim 1, characterized in that: The safety production education and training model is used to implement intelligent dialogue learning for the target users, and a personalized education and training program is generated for each target user, specifically including: Obtain user profile information and training course information of the target user; the user profile information of the target user includes at least the basic identity information, historical training records and historical evaluation results of the target user; the training course information includes at least the course introduction, difficulty level, training time and knowledge points; Automatically generate question guidance information based on the user portrait information of the target user; the question guidance information at least includes ToolboxTalk mode prompt words; Collecting questions or needs raised by the target user according to the question guiding information, and determining personalized information of the target user based on the questions or needs; the personalized information at least includes the course preference of the target user; Extracting a first feature vector corresponding to the user portrait information of the target user and a second feature vector corresponding to the personalized information; Based on the first feature vector and the second feature vector, the dialogue strategy and recommended content are adjusted based on the safety production education and training model dynamically combined with collaborative filtering or content filtering methods to generate a personalized education and training plan.

6. The intelligent training method for production safety education according to claim 5 is characterized in that: The method of using the production safety education and training model to implement intelligent dialogue learning for the target users and generating a personalized education and training program for each target user also includes: Performing sentiment analysis on the training course information to identify sentiment tags corresponding to the training course information; Using a support vector machine combined with the emotional tags to perform emotional classification on the training course information, and constructing an emotional database; Generate personalized multimedia education and training content for the target user, and dynamically adjust the multimedia education and training content in combination with the emotion database; the multimedia education and training content includes at least animation, video, text, audio and picture; the dynamic adjustment includes at least adjusting the tone, speech speed, sound effects, background music, animation character expression or action.

7. The intelligent training method for production safety education according to claim 6, characterized in that: After generating personalized multimedia education and training content for the target user, the method further includes: Acquire interactive behavior information between the target user and the multimedia education and training content; the interactive behavior information at least includes click information, comment information and viewing time of the target user on the multimedia education and training content; Based on the interactive behavior information and combined with reinforcement learning, the multimedia education and training content is dynamically adjusted; the multimedia education and training content supports exporting and sharing in multiple media formats.

8. The intelligent training method for production safety education according to claim 7 is characterized in that: The automatic generation of an assessment question bank according to the personalized education and training program for each target user specifically includes: Automatically generate assessment questions based on the production safety knowledge base, the training course information and pre-stored question templates; Based on the target user's learning progress for the personalized education and training program and the interactive behavior information, automatically adjust the difficulty of the assessment questions and generate an assessment question bank for the target user; After generating the test question bank for the target user, the method further includes: Obtaining the target user's assessment score for the assessment question bank, performing intelligent assessment for the target user, obtaining the assessment result, and collecting feedback data from the target user; Based on the evaluation results and the feedback data, the production safety education and training model is updated.

9. An intelligent training system for production safety education, characterized in that the device include: Safety production knowledge base, personalized education and training program generation module, and assessment question bank generation module; The safety production knowledge base is constructed from basic data related to safety production; The personalized education and training program generation module builds a safety production education and training model based on the safety production knowledge base and user portrait data; and uses the safety production education and training model to implement intelligent dialogue learning for target users, and generates a personalized education and training program for each target user; the personalized education and training program at least includes multimedia education and training content; the safety production related basic data at least includes industry standard procedure data and accident case data for various types of work and various positions in the safety production industry; The user portrait data at least includes basic user identity data of the target user; The assessment question bank generation module is used to automatically generate an assessment question bank according to the personalized education and training program of each target user.

10. An intelligent training device for production safety education, characterized in that the device include: A memory, a processor, and a communication interface coupled to the processor; The memory stores a computer program executable by the processor; When the processor runs the computer program, it executes the intelligent training method for production safety education as described in any one of claims 1 to 8.

Images