The Informa ionization of social life and the globalization of economy have made the importance of English increasingly prominent. Building an information-based teaching platform for supplementary teaching under the network environment has become a mainstream teaching method in various basic schools. How to integrate various types of multimedia teaching resources into English classroom teaching has become the main goal of the current college teaching reform. Aiming at the shortcomings of the current English education classroom, this paper designs and develops an English education cloud classroom based on the Internet of Things and data mining methods. First of all, the system adopts a three-tier B/S model, the development platform chooses, NET, the development language, uses ASP.NET, and the database chooses SQL server. Secondly, the data mining method is used to clean and organize the data in the cloud classroom background to explore the course education status behind the data. Finally, the simulation test analysis verifies the efficiency of the English education cloud classroom established in this article.
Institutions of higher learning are the specific implementation link of our country’s talent strategy. In recent years, our country has vigorously promoted educational reform and educational technology application [
The use of big data technology can extract laws from massive data [
From the perspective of the functionality of the online teaching platform, the content of the platform is more static, and its dynamic personalized recommendation and evaluation functions are not yet mature [
With the gradual deepening of distance education research, English distance education researchers have increasingly realized that comprehensive and systematic learning support services are the core element of maintaining the success of English distance education [
Schematic diagram of english education cloud classroom concept.
With the gradual deepening of research on English distance education, English distance education researchers have increasingly realized that comprehensive and systematic learning support services are the core elements to maintain the success of English distance education [
In recent years, mobile information technology has received extensive attention and development. Technologies such as edge computing and the Internet of Things have received more and more attention from researchers. Definition of the Internet of Things: the Internet of Things (IoT) can be regarded as a far-reaching vision with technical and social significance. From the perspective of technology standardization, IoT can be regarded as the infrastructure of theglobal information society, providing physical interconnection (physical and virtual) on the basis of existing and emerging interoperable information and communication technologies (ICT) advanced business. Through identification, data capture, processing, and communication capabilities, IoT can make full use of “things” to provide services for various applications, while ensuring security and privacy requirements [
The prevailing Internet of Things platform is essentially a centralized structure. Although the traditional Internet of Things is trying to use emerging distributed storage, edge computing, and other technologies, it has not changed its centralized nature [
Data mining methods include machine learning methods, statistical methods, neural network methods, and database methods. Among them, machine learning methods include inductive analysis methods (decision trees, rule induction, etc.) and genetic algorithms. Statistical methods include regression analysis (autoregressive, multiple regression, etc.), discriminant analysis (Bayesian discriminant, Fischer discriminant, and nonparametric discriminant), and cluster analysis. Neural network methods include feedforward neural network (BP algorithm) and self-organizing neural network. Database methods include multidimensional data analysis and OLAP methods. Among these methods, the following mining algorithms are commonly used: decision trees, association rules, Bayes, neural networks, rule learning, etc.
Gradient Boosting Decision Tree (GBDT) is a classic boosting algorithm. It is based on the idea of boosting algorithm, and in each iteration, a new decision tree is established in the direction of reducing the gradient of the residual and iteratively improves the generalization ability of the system. The gradient boosting decision tree is essentially a combination of multiple decision trees. The decision tree algorithm based on gradient boosting can identify distinguishable features and feature combinations. In the GBDT algorithm, the path of the decision tree can be directly used as the input features of other models, reducing the steps of manually selecting and combining features. Therefore, in the context attribute weight calculation, it is possible to identify context attributes that affect user preferences and to obtain the weight results of context attributes based on the relationship between context attributes, so as to dig deeper into user needs and provide users with more personalized Information recommendation. Conceptual diagram of decision tree data mining is shown in Figure
Conceptual diagram of decision tree data mining.
Discretize each context instance under context attributes, convert them into input features, and input them into the gradient boosting decision tree. With the advent of the era of big data and artificial intelligence, the learning methods, teaching methods, and cognitive methods of English distance education have undergone major changes. The characteristics of the era require new connotations to be injected into learning support services; from the traditional unified, the fixed learning support services have shifted to the development of personalized English teaching design, curriculum management, and learning evaluation services. Since the gradient boosting decision tree algorithm is composed of multiple decision trees, each decision tree uses a top-down greedy algorithm to select the attribute with the best classification effect at each node to split:
Therefore, the reference documents of this study measure the contribution degree of the situation instance to the user’s choice based on the average change of the Gini index when each situation instance is used as a split node in each decision tree:
Assume that M decision trees are obtained through the GBDT algorithm according to the user’s preference information for selecting information resources and then the context instance Ck under the context attribute. The degree of contribution to the user's choice of information resources originates from the situational instance Ck:
When M is used as a split node in a decision tree, there is no change in the average value of the Gini coefficient. Among them, the calculation formula for the Gini index of node and node is
The priority of each attribute of the decision tree is usually based on the information gain:
In addition, the Gini index and gain ratio are also commonly used to divide optimal attributes, where the gain ratio is expressed as
In order to prevent overfitting, decision trees usually adopt pruning methods to improve generalization:
Pruning is divided into prepruning and postpruning. Prepruning is based on the calculation result of information gain to determine in advance whether retaining nodes will increase the generalization of the model. Postpruning is to first generate a complete decision tree model and then proceed upward from the bottom leaf node. Investigate and decide whether to keep each node. For the situation where a sample can belong to multiple categories at the same time, the existence of the degree of membership is used to reflect the degree to which the sample belongs to a certain category. Fuzzy mathematics can express the fuzzy nature of things and relationships. On this basis, a fuzzy fault diagnosis model can be constructed to enable fault diagnosis to better handle the complex relationship between fault sources and fault symptoms.
The cloud classroom for English education based on data mining is an aid and extension of classroom teaching and is a tool to help students achieve after-class review and consolidate and reduce the workload of teachers. The biggest feature of the system should be reflected in individualization, that is, according to the characteristics of students, the information obtained by data mining should be used to dynamically select and organize the materials to be learned in teaching resources so that students can learn. Individualized guidance can be obtained in the selection of content, the understanding of learning goals, the evaluation of learning effects, and the diagnosis of the learning process, so as to truly realize teaching in accordance with their aptitude. The framework design of the English education cloud classroom system is shown in Figure
Framework design of the English education cloud classroom system.
The functional roles of the English education cloud classroom system can be divided into three types according to the user’s authority, including system administrators, teachers, and students. For the situation where a sample can belong to multiple categories at the same time, the existence of the degree of membership is used to reflect the degree to which the sample belongs to a certain category. Fuzzy mathematics can express the fuzzy nature of things and relationships. On this basis, a fuzzy fault diagnosis model can be constructed to enable fault diagnosis to better handle the complex relationship between fault sources and fault symptoms. The responsibilities of the system administrator include the maintenance and management of system information, user information, and user rights; the responsibilities of teachers include the management of teaching resources, through data mining of English student information to analyze and evaluate students’ learning behaviors and adjust teaching strategies; students are in the system’s personalized learning interface for autonomous learning, practice, testing, and answering questions.
In order to verify the impact of the English education cloud classroom system on the learning effect of students, this research applies the system to the undergraduate English exam tutoring course. The effectiveness of the system is tested through three methods: the pass rate of the English unified test, teacher interviews, and student questionnaires, and the application effects of the system are analyzed to find the direction for improvement in the later period. Distribution of score data in English education cloud classroom is shown in Figure
Distribution of score data in English education cloud classroom.
Figure
The number distribution of classrooms in different situations.
Figure
Comparison of english classroom based on gradient boosting decision tree and other methods.
Figure
From the current technical point of view and teaching task requirements, it is not feasible to completely use the Internet teaching platform to replace manual teaching. However, it is feasible to use the Internet teaching platform as an auxiliary means of manual teaching or even as a teaching platform for elective courses. With the rapid development of network technology, modern teaching has an increasingly urgent need for a mature network teaching platform. Based on this demand, it is necessary to develop and design network teaching systems and gradually apply them to teaching activities. It is necessary to improve the quality of education of. Combining the characteristics of autonomous learning and collaborative learning, comprehensively considering the needs of teaching management, resource sharing, and multidirectional interaction, this article creates a new type of English education network teaching system that conforms to the 21st century education information construction. The teaching system adopts the three-tier architecture of B/S in the choice of architecture. This choice can make the system’s operating ability more improved and run more smoothly. With the advancement of information technology and network technology, more and more educators realize the importance of network teaching, and more and more network teaching systems are applied to teaching. Therefore, a new network teaching system that fully meets the needs of teachers and students can achieve more long-term development.
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
This article does not contain any studies with human participants or animals.
All authors agree to submit this version and claim that no part of this manuscript has been published or submitted elsewhere.
The author declares that he has no conflicts of interest.