A Hybrid Teaching System for Vocational College Clothing Major using Bidirectional and Auto-Regressive Transformers

Now a days, the advent of digital technology has transformed the education sector from revolutionized traditional teaching to learning methods. This technological integrates has significantly impacted education system that leading to the development of a pedagogical approaches and enhanced student engagement. However, this technology leads to the lack of critical thinking and problem-solving skills, which is crucial for students' emotional and mental well-being. To overcome those drawbacks this paper, proposed a recommendation system based on the Deep Learning (DL). Initially, data is collected from the Tex atlas dataset, serving as the foundation for the hybrid teaching system for Vocational College Clothing. Next, a Vocational Auto Encoder is employed to extract optimal attributes from the collected data, distilling the most relevant information. Moreover, the refined data is utilized to generate new data through a robust model of Bidirectional and Auto-Regressive Transformers (BART). This proposed model leverages the extracted attributes to produce high-quality with synthetic data that can be used for various applications. Moreover, by combining these cutting-edge techniques, the process capable of producing valuable insights and augmenting existing datasets. The performance of proposed model evaluated with the various metrics and achieved accuracy of 93.67%, specificity of 94.56 %, recall of 95.26 % and precision of 93.28% by comparing with the existing Bidirectional Encoder Representations from Transformers (BERT) method on recommendation system.