Intelligent education consultation management system based on adaptive learning path recommendation

Abstract

The invention provides an intelligent education consultation management system based on adaptive learning path recommendation, and relates to the technical field of computer science, and the system comprises a data collection module, a network construction module, a model construction module and a deployment module. The system constructs an adaptive learning path recommendation and emotion recognition model by collecting interaction data and emotion information between teachers and students, thereby recommending learning paths and learning resources in a personalized manner. According to the system, the artificial intelligence technology and education domain knowledge are combined, customized learning support is provided for each student, and the participation degree of the students and reasonable configuration of education resources are improved.

Description

Technical Field

[0001] The present invention relates to the field of computer science and technology, and in particular to an intelligent education consulting management system based on adaptive learning path recommendation. Background Art

[0002] With the rapid development of artificial intelligence technology, more and more online communication and interaction platforms have emerged, such as online chat rooms, instant messaging applications and social networking platforms. On these platforms, people can freely communicate and exchange in various forms such as text, voice and video anytime and anywhere. The emergence of these platforms has provided more convenience and flexibility for people's communication, while also creating more job opportunities and development space, promoting economic prosperity and social development.

[0003] However, among the many online communication and interaction platforms, the online dialogue consultation system in the field of education is a very important part with broad development prospects. In recent years, with the popularization of Internet technology and big data applications and the increasing attention paid to education by the country, more and more people have begun to use these emerging network tools to seek help and learn knowledge. Therefore, how to establish an efficient, intelligent and high-quality education dialogue consultation system is one of the problems that need to be solved urgently. At present, most of the existing education dialogue consultation systems have relatively single functions and can only rely on manual intervention to complete students' academic planning and guidance work; in addition, due to the lack of sufficient auxiliary analysis and prediction methods, it is difficult to achieve the goals of precision and high efficiency. If we want to truly give play to the advantages of these platforms, we need to design a more advanced and effective intelligent service system to meet the needs of students with different needs. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide an intelligent education consulting management system based on adaptive learning path recommendation, which can solve the problems of insufficient personalized teaching support, low student participation and uneven distribution of educational resources in online dialogue consulting systems in the field of education.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] An intelligent education consulting management system based on adaptive learning path recommendation, comprising:

[0007] The data collection module is used to collect information about the interaction between teachers and students to obtain the dialogue text data used in the education field, and to collect students' emotional vocabulary and expression pictures in the interaction to build a large-scale emotional dictionary, and to classify and annotate them to obtain emotional analysis data;

[0008] A network construction module, used to construct a neural network structure for automatic recommendation of learning paths using the conversation text data;

[0009] The first model building module is used to design a corresponding loss function to measure the performance of the neural network structure, and use an optimization algorithm to train the neural network structure to obtain a trained adaptive learning path recommendation model; the adaptive learning path recommendation model is used to recommend a personalized learning path based on the student's dialogue text data;

[0010] A second model building module is used to establish an emotion recognition model based on the emotion analysis data; the emotion recognition model is used to extract key emotion features from the conversation and push relevant learning resources based on the emotion features;

[0011] A deployment module is used to deploy the adaptive learning path recommendation model and the emotion recognition model on an online education platform.

[0012] Preferably, the dialogue text data includes: question and answer records between teachers and students, discussion summaries, and related background information and parameters.

[0013] Preferably, the optimization algorithm is a gradient descent method.

[0014] Preferably, it also includes:

[0015] The data sorting module is used to perform data preprocessing on the conversation text data; the data preprocessing includes: data cleaning, data labeling and data encoding.

[0016] Preferably, the data cleaning includes: deleting duplicate or incomplete information and filtering invalid information.

[0017] Preferably, the data preprocessing further includes: data conversion; the data conversion includes:

[0018] Use machine translation tools to convert the original text into a preset language;

[0019] Transcribe the audio file into written form.

[0020] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0021] The present invention provides an intelligent education consultation management system based on adaptive learning path recommendation, including: a data collection module, which is used to collect information on the interaction between teachers and students to obtain the dialogue text data used in the education field, and collect the emotional vocabulary and expression pictures of students in the interaction to build a large-scale emotional dictionary, and classify and label to obtain emotional analysis data; a network construction module, which is used to use the dialogue text data to build a neural network structure for automatic recommendation of learning paths; a first model construction module, which is used to design a corresponding loss function to measure the performance of the neural network structure, and use an optimization algorithm to train the neural network structure to obtain a trained adaptive learning path recommendation model; the adaptive learning path recommendation model is used to recommend personalized learning paths according to the students' dialogue text data; a second model construction module, which is used to establish an emotional recognition model according to the emotional analysis data; the emotional recognition model is used to extract key emotional features from the dialogue, and push relevant learning resources according to the emotional features; a deployment module, which is used to deploy the adaptive learning path recommendation model and the emotional recognition model on an online education platform. The present invention combines artificial intelligence technology with knowledge in the field of education, provides personalized learning experience and support for each student, improves student participation and also realizes the reasonable allocation of educational resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0023] Figure 1 A schematic diagram of the system structure provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] Figure 1 A schematic diagram of the system structure provided by an embodiment of the present invention, such as Figure 1As shown, the present invention provides an intelligent education consultation management system based on adaptive learning path recommendation, comprising:

[0027] The data collection module is used to collect information about the interaction between teachers and students to obtain the dialogue text data used in the education field, and to collect students' emotional vocabulary and expression pictures in the interaction to build a large-scale emotional dictionary, and to classify and annotate them to obtain emotional analysis data;

[0028] A network construction module, used to construct a neural network structure for automatic recommendation of learning paths using the conversation text data;

[0029] The first model building module is used to design a corresponding loss function to measure the performance of the neural network structure, and use an optimization algorithm to train the neural network structure to obtain a trained adaptive learning path recommendation model; the adaptive learning path recommendation model is used to recommend a personalized learning path based on the student's dialogue text data;

[0030] A second model building module is used to establish an emotion recognition model based on the emotion analysis data; the emotion recognition model is used to extract key emotion features from the conversation and push relevant learning resources based on the emotion features;

[0031] A deployment module is used to deploy the adaptive learning path recommendation model and the emotion recognition model on an online education platform.

[0032] Specifically, this embodiment provides an offline training method, the purpose of which is to improve the prediction accuracy and versatility of the system. The offline training method includes the following main stages:

[0033] 1. Educational dialogue data collection: collect information about the interaction between teachers and students on the online educational interactive platform, and obtain dialogue text data used in the educational environment. These data include question and answer records, discussion summaries, and related background information and parameters between teachers and students.

[0034] 2. Training of the adaptive learning path recommendation model: Using the obtained educational dialogue dataset, develop a neural network structure for automatic learning path recommendation. In this process, it is necessary to design a corresponding loss function to measure the performance of the model. Use the gradient descent method or other optimization algorithms to train the model.

[0035] 3. Training of sentiment analysis model: Collect students’ sentiment words and expression pictures when they interact with teachers and other peers, build a large-scale sentiment dictionary and classify and annotate them. Then build a sentiment recognition model to extract key sentiment features from the conversation.

[0036] In the second aspect, the present invention also provides an online application method based on adaptive learning path recommendation and sentiment analysis, which aims to use artificial intelligence technology and advanced machine learning models to provide each student with personalized learning resources and suggestions, and adjust teaching strategies in real time to promote their learning and growth. The online application method has the following main functions:

[0037] 1. User login and personal data management: Students can register and log in through the education consulting system, and the system will automatically collect and save the user's personal information (such as name, age, gender, subject, etc.). In addition, user preferences can also be set, such as the type of study and learning difficulty they prefer.

[0038] 2. Education consulting system interaction: Students can search, browse or search for topics they are interested in in the system. When they find a topic they want to learn, they can start communicating with experts / teaching assistants (i.e., the system) on the topic and ask and answer questions or discuss the content of a topic with each other through natural language.

[0039] 3. Adaptive learning path recommendation: After each communication, the system will analyze the current chat context based on previous historical data, and recommend the best task list for the current scenario to the student. This recommendation not only takes into account the student's previous performance and preferences, but also the situational factors at the time. The ultimate goal is to help every student get the most appropriate and efficient learning experience.

[0040] 4. Real-time feedback and learning effect tracking: The system provides a timely evaluation feedback mechanism, allowing students to give their own opinions and evaluations immediately after completing a certain stage of the course. This will make the entire learning process more transparent and visual. At the same time, the system will continue to monitor the performance of each student and compare and evaluate it with the preset standards. If any abnormal situation is found, the early warning program will be immediately activated to ensure timely intervention and problem handling.

[0041] 5. Emotion-driven teaching strategy adjustment: During the entire conversation, we also need to pay attention to the trend of students' emotions and make adjustments when necessary. We can first perform a preliminary analysis by emotional classification of text data (i.e., emotional labeling), and then use speech signal processing tools to further detect whether the speaker's words match the situation in which he is in harmony. If it is determined that there are some problems, special measures can be considered to ease the tension and ensure the feasibility and reliability of the entire communication channel.

[0042] This embodiment provides an implementation scheme of an intelligent education consulting system for adaptive learning paths and emotion recognition. The scheme uses the latest advances in artificial intelligence technology to promote the goals of personalized education and improving student engagement. Specifically, the system includes the following main components and functions:

[0043] (a) Educational dialogue dataset based on big data;

[0044] (b) Construction of personalized and customized knowledge base based on adaptive learning;

[0045] (c) Sentiment analysis and emotion monitoring module;

[0046] (d) Online education consultation interactive platform.

[0047] Among them, (a) and (b) are offline training processes, while the other three components are online application processes. During the training process, a large amount of educational dialogue data needs to be collected and processed to build a comprehensive system with adaptive learning and sentiment analysis capabilities. In actual applications, the system can be used as part of an online dialogue platform to provide personalized educational services for students.

[0048] (a)(i): Collection of educational conversation data. First, for experts in the field of education, they often participate in the process of educational tutoring, such as teaching mathematics or physics to high school or college students. These experts may generate some data files of records or voice recordings when teaching, such as course handouts on geometry topics and voice recordings on vector differentials. The raw audio and text information generated in this way is regarded as "educational conversation data". Secondly, in addition to advanced training on specific topics, ordinary students often seek help and support from teachers online. Therefore, we can obtain a large amount of data reflecting daily school interactions from various online platforms. These data mainly include two aspects: one is unstructured text chat messages posted by users from different platforms, and the other is sound clips during video calls. Since this type of data has rich corpus and contains real teacher-student communication scenes, we call it one of the important components of "educational conversation data". Finally, considering that educational conversations often only involve one-to-one teacher-student communication modes, we need to further expand the scope of the database to cover more types of conversation types. For example, if a student introduces himself and asks some questions to his teacher before talking to him, then the relevant text materials should be retained in this case. In addition, if there are many discussion groups on the Internet about how to solve problems or other related issues, they should also be included in our database as much as possible, because they are also a valuable source of educational dialogue materials. In short, all educational content that has happened online should be classified under this broad scope to ensure that all necessary materials for the successful completion of the entire project are fully guaranteed. It is worth noting that since we want to create a very flexible and scalable platform architecture to cope with possible changes and changes in technical requirements in the future, it is impossible to fully cover all possibilities in the initial stage, but at least it can be gradually established in the above way. And as time goes by and we have a deeper understanding of the actual situation, this system will be gradually enriched and improved.

[0049] (a)(ii): Preprocessing of conversation data. After we have obtained enough conversation data for subsequent research, we need to perform preliminary processing on these original materials before we can truly analyze and study them. Specifically, there are several operation processes to choose from: (i) delete duplicate or incomplete information; (ii) filter out obviously irrelevant parts; (iii) use machine translation tools to convert the original text into a language that is more convenient to read; (iv) transcribe the audio file into written form, etc. Only after these preliminary preparations can we carry out the next step of research more effectively, otherwise it will bring unpredictable risk factors to the results. In addition, there are several precautions that need to be mentioned here. The first is to ensure that the selected processing procedures must comply with international standards and relevant laws and regulations; the second is to ensure that each link can be traced back to the origin without causing any errors or mistakes; and then it is best not to rely too much on manual intervention but to rely more on automated means to achieve the corresponding operations, which can greatly save time and cost. In addition, it should be noted that strict testing standards need to be set for each link to ensure that the overall quality is not affected. The final results need to be submitted to professionals for review to assess their accuracy and reliability before deciding whether to continue using them.

[0050] (b)(i): Establishment of adaptive learning model. This section focuses on how to develop an adaptive learning model suitable for specific situations based on existing educational dialogue data, so that students can enjoy more personalized services and better learning experience during the learning process.

[0051] The following points need to be explained:

[0052] (I) The model is mainly composed of a two-layer neural network and trained through deep reinforcement learning;

[0053] (II) It can automatically search for the best matching path in a short time and give corresponding suggested guidance directions;

[0054] (III) A variety of variable parameters are considered before each decision to ensure the most accurate effect output;

[0055] (IV) During actual operation, certain error correction mechanisms can also be used to prevent problems such as system crashes caused by unexpected situations.

[0056] (b)(ii): Basic principles of model training. Here is a brief introduction to how the model works in a real environment. In actual application, whenever we receive a new request, we will start the entire intelligent control system and analyze and judge the problem to be solved. After confirming that the solution is correct, we enter the execution phase and multiple sub-nodes work together to achieve the expected goal. In the whole process, the intelligent terminal based on previous experience plays a key role - responsible for receiving instructions and issuing commands and real-time monitoring of the changes in the values ​​of various indicators. At the same time, other auxiliary equipment can also play a supplementary role to a certain extent. For example, using sensors to detect information such as temperature and humidity in the surrounding environment, and capturing image information through cameras, etc., are important measures taken to make the entire system more accurate and reliable. In comparison, the disadvantages of the traditional programming method mentioned in the conventional practice are much more insignificant. In fact, the latter not only relies too much on human subjective consciousness, making it difficult to accurately grasp certain problems, but also easily leads to the emergence of a situation of loss of control when facing emergencies.

[0057] (c)(i): Basic concepts and classification of emotion recognition. Emotion refers to people's attitude towards the outside world. It is both a feeling and a complex multidimensional psychological phenomenon, including but not limited to various forms of expression such as joy, anger, sorrow and happiness. Generally speaking, we often encounter various emotional fluctuations in our daily lives, and these emotional changes are usually closely related to a person's specific behavioral characteristics. For example, when people hear unfortunate news or see touching scenes, they will feel sad, sad, or even desperate and helpless, and other negative emotions; however, when facing challenges and difficulties, they will show positive mental reactions such as courage, confidence, and even enthusiasm. It can be seen that various special experiences generated in different occasions can reflect the objective situation of the individual at this stage and point out the direction for future development. However, as a complete cognitive process, it is not simple and easy, because it is also affected by many other additional factors such as genetic background, growth experience and living environment. This means that even if the same individual may experience the same event under the same conditions, he may show completely opposite emotional changes. For example, if we see the same painting in the same place, but each has a completely different interpretation angle and understanding method, resulting in two extremely different evaluation conclusions. The same principle also exists in interpersonal relationships. Because everyone has different personality traits and life experiences and backgrounds, they may have different reactions and thinking tendencies when faced with the same thing.

[0058] (c)(ii): Emotion recognition method based on natural language. This method uses computer programs to simulate the analytical thinking process of the human brain and designs corresponding calculation rules on this basis to assist us in tasks such as judging and reasoning about real things. It can be applied not only to fields such as medical psychology, but also to many social hot topics such as news reports and film and television reviews. The most typical representative case is the image display of the "Cortana" robot product launched by Microsoft. Through it, it can easily and happily have multiple rounds of conversations with ordinary people and correctly perceive the other party's intentions and actively respond to opinions. In addition, it has also been fully reflected in the film industry. For example, in the "Artificial Intelligence" series of films, there are specially designed virtual assistant characters and their owners who use oral methods to engage in lively, interesting and intelligent dialogue scenes. Of course, similar applications can also be found in other entertainment programs and game works. In summary, the theoretical basis of this method comes from the knowledge reserves of related disciplines such as psychology and social sciences. At the same time, it has also benefited from the rapid development of computer technology and big data analysis technology in recent years. It has become one of the technological innovations that is highly practical and easy to promote and popularize. In general, by making rational use of this method, people's attention and sensitivity to things around them can be effectively enhanced, thereby enabling them to demonstrate more outstanding work ability and leadership talent in their daily life and work.

[0059] (d): Online application of intelligent education consulting system. This is one of the core functional modules of this system. It can integrate all the technical solutions mentioned above into a unified whole and maximize their benefits. In the actual work process, we can interactively control it through the user interface or other forms to achieve all the requirements of online education consulting. At the same time, this process will also be accompanied by the synergy of a series of related technologies, such as data analysis statistics, information retrieval and screening functions. In this way, users can enjoy a more comfortable and convenient service experience. The specific implementation method of the present invention also includes the following: - Collection and collation of educational dialogue data. The dialogue data between students and teachers is collected through the online education platform, and the data is cleaned, labeled and encoded to form a high-quality data set. - Training of adaptive learning path recommendation model. Using the trained data set, the model is trained and tuned so that the model can be dynamically adjusted according to the student's learning situation and recommend appropriate and personalized learning paths. - Training of sentiment analysis model. Combined with the student's dialogue information and sentiment tags, the model is trained and tuned so that the model can identify the student's emotional state and push suitable related learning resources to it. - Online intelligent education consulting system. The trained model is deployed on the online education platform to answer and guide students' questions, helping them better understand and master knowledge.

[0060] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0061] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. An intelligent education consulting management system based on adaptive learning path recommendation, characterized in that: include: The data collection module is used to collect information about the interaction between teachers and students to obtain the dialogue text data used in the education field, and to collect students' emotional vocabulary and expression pictures in the interaction to build a large-scale emotional dictionary, and to classify and annotate them to obtain emotional analysis data; A network construction module, used to construct a neural network structure for automatic recommendation of learning paths using the conversation text data; The first model building module is used to design a corresponding loss function to measure the performance of the neural network structure, and use an optimization algorithm to train the neural network structure to obtain a trained adaptive learning path recommendation model; the adaptive learning path recommendation model is used to recommend a personalized learning path based on the student's dialogue text data; A second model building module is used to build a sentiment recognition model based on the sentiment analysis data; The emotion recognition model is used to extract key emotion features from the conversation and push relevant learning resources based on the emotion features; A deployment module is used to deploy the adaptive learning path recommendation model and the emotion recognition model on an online education platform.

2. The intelligent education consulting management system based on adaptive learning path recommendation according to claim 1 is characterized in that: The dialogue text data includes: question and answer records between teachers and students, discussion summaries, and related background information and parameters.

3. The intelligent education consulting management system based on adaptive learning path recommendation according to claim 1 is characterized in that: The optimization algorithm is the gradient descent method.

4. The intelligent education consulting management system based on adaptive learning path recommendation according to claim 1, characterized in that: Also includes: A data sorting module, used for performing data preprocessing on the conversation text data; The data preprocessing includes: data cleaning, data labeling and data encoding.

5. The intelligent education consulting management system based on adaptive learning path recommendation according to claim 4 is characterized in that: The data cleaning includes: deleting duplicate or incomplete information and filtering invalid information.

6. The intelligent education consulting management system based on adaptive learning path recommendation according to claim 4, characterized in that: The data preprocessing also includes: data conversion; the data conversion includes: Use machine translation tools to convert the original text into a preset language; Transcribe the audio file into written form.

Images