Effectiveness of Outcome-Based Education for Student Performance Empowerment in Tertiary Education System based on Split RNN Approach

Institutional and systemic changes are occurring at a dizzying rate in the world of higher education. In light of larger difficulties that have persisted since the new millennium began, this research looks at how things are in higher education now, including new developments and structural changes. There is a pressing need for flexible approaches to education in light of the ways in which students, training requirements, institutional frameworks, leadership, and economic development have been influenced by the information and communication revolution and globalisation. Data preprocessing, feature extraction, and model training comprise the structured machine learning approach that was created to improve student performance and empowerment in the tertiary education system. For feature extraction, the annotated TF-IDF method was utilised, after preprocessing methods like stemming and stop-word deletion. To boost the model's predicted accuracy, a Split BiRNN was used during training. With a success rate of 97.48%, the suggested method surpassed traditional models such as RNN and BiLSTM. These results demonstrate that cutting-edge machine learning methods are useful for improving academic outcomes and school policy. This research empowers students and improves educational results by tackling structural obstacles and incorporating data-driven solutions, contributing to the continual evolution of the tertiary education system.