Online Teaching System using Linformer based Attention for Scalable Educational Sequence Modeling

As digital learning environments become increasingly central to modern education, the need for intelligent and adaptive online teaching systems has grown significantly. Traditional systems often lack the capacity to process and interpret large-scale, unstructured teaching data efficiently, limiting their ability to support dynamic content delivery and real-time interaction. To overcome these limitations, this research introduces a deep learning (DL) based online teaching framework built on the Linformer architecture. Linformer, a streamlined variant of the Transformer model, is leveraged for its ability to handle long-sequence educational data with reduced computational overhead. A comprehensive dataset comprising digital classroom activities, teacher-student interactions, and multimedia teaching content was used to train and evaluate the system. Experimental results show that the proposed model achieves a high accuracy of 98.5% in aligning instructional content efficiently. Furthermore, performance indicators including Precision (97.1%), Recall (97.8%), F1-Score (97.4%), and Specificity (98.2%) confirm the model’s strong adaptability and reliability in managing diverse teaching scenarios. This study demonstrates the potential of efficient attention mechanisms in enabling scalable, responsive, and intelligent online teaching systems, paving the way for more interactive and personalized digital instruction.