Application of mobile learning system based on convolutional network technology in students' open teaching strategies

Area of Science:

• Artificial Intelligence
• Educational Technology
• Machine Learning

Background:

• Open teaching strategies require effective personalized learning resource recommendations.
• Mobile learning systems need intelligent support for diverse user needs.
• Existing recommendation systems often lack personalization and adaptability.

Purpose of the Study:

• To design and develop a mobile learning system utilizing a convolutional neural network.
• To propose an advanced TCN-DCC-RL model for personalized learning resource recommendations.
• To evaluate the model's effectiveness using real-world learning analytics data.

Main Methods:

• Development of a mobile learning system integrated with a convolutional neural network.
• Implementation of a Temporal Convolutional Network (TCN) with Dilated Causal Convolution (DCC) and Reinforcement Learning (RL).
• Evaluation of the TCN-DCC-RL model on the UK Open University's learning analytics dataset.

Main Results:

• The TCN-DCC-RL model demonstrated high performance metrics: 96.49% accuracy, 90.57% F1-score, 92.27% Mean Average Precision, and 0.938 NDCG@20.
• Significant enhancement in the personalization and intelligence of learning resource recommendations.
• Validation of the model's effectiveness in a practical educational context.

Conclusions:

• The proposed TCN-DCC-RL model offers a novel technical approach for intelligent education systems.
• The system effectively supports open teaching strategies through personalized recommendations.
• Future intelligent education systems can benefit from this advanced personalization technique.