Application of Artificial Intelligence Technology in Adaptive System for English Speaking Learning

Abstract

There are still issues, though, such as data bias that can result in erroneous assessments, the possibility of relying too much on technology at the expense of conventional techniques, and privacy issues brought on by data collection. Careful navigation is also required due to the limitations of AI algorithms in understanding contextual nuances and the ethical issues surrounding its use. Even though it shows promise, resolving these issues is essential to optimizing the efficacy and moral consequences of AI-based English learning. This manuscript proposes Application of Artificial Intelligence Technology in Adaptive System for English Speaking Learning (AAIT-ASESL-MORA-RNN). The input data is gleaned from the English Language Learning-Ensemble Learning dataset. Then, the data is provided to preprocessing phase. During the pre-processing phase, the Unscented Trainable Kalman Filter (UTKF) is used to identifying the missing data’s.  Then the preprocessed data are fed to the Mixed-Order Relation-Aware Recurrent Neural Networks (MORARNN) and predict the Error of English Speaking Learning.  In general, MORARNN does not express adapting optimization strategies to determine optimal parameters using a MORARNN. The MORARNN is optimized using the Hermit Crab Optimizer (HCO). The proposed method is implemented in Python. The proficiency of the AAIT-ASESL-MORA-RNN approach is evaluated using a number of performance criteria, including accuracy, recall, precision, F1-Score. The proposed AAIT-ASESL-MORA-RNN method covers 28.36%, 23.42% and 33.27% higher precision and of 22.36%, 15.42% and 18.27% higher accuracy compared with existing AI basis English Self-Learning Effect Evaluation with Adaptive Influencing Factors Analysis (ESL-AIF-AI) Research on English Hybrid Assisted Teaching Method under Contextual Support of R-CNN (EHATS-CS-RCNN) and College English Teaching Quality Evaluation Scheme depending on Information Fusion along Optimized Radial Basis Function Neural Network (CET-QES-RBFNN) respectively.