Automatic English Translation using Artificial Recurrent Neural Network with Machine Translation

In recent years, automated recognition technique for English translation error recognition yields poor accuracy recognition results due to its limited ability for semantic analysis. Consequently, this research creates an automatic identification approach for Machine Translation (MT) error detection based on Artificial Intelligence (AI). Thus, this research developed an intelligent identification method known as Artificial Recurrent Neural Network with Machine Translation (ARNN-MT) in order to recognize English translation errors. Context-learning from previous words is used by ARNN sequential data to detect translation issues. MT uses these networks to predict accurate translations. These hybridized ARNN-MT techniques combined with rule-based process, and learned patterns to enhance error detection. Consequently, ARNN-MT improves phrase flow and able to resolve specific translation problems while providing a logical translation in context. When compared to existing algorithms, including Machine English Translation Errors-Multifeatured Fusion (METE-MF), Double-Recurrent Neural Network (Double RNN), and Radio Magnetic Pronunciation Recording Devices (RMPRD), the results demonstrated that the proposed ANN-MT method performed better in terms of accuracy, of 99.96%.