A Method for Evaluating English Translation Quality of Intelligent Translation Robots based on Long Short-Term Memory

English is a significant languages for economic conversation in different countries among worldwide which is extensive used language for data exchange. China has high importance in English language and individuals demand for learning is hastily in nowadays. This work proposed a Deep Learning (DL) based technique named as Long Short-Term Memory (LSTM) for evaluating English translation quality of intelligent translation. The English translation function is premeditated and entire module is calculated which realizes data acquisition, processing and output. The data collection is applied to assemble speech and audio input device is utilized for English signal to processing scheme for processing data signal. The processed outputs are respective to client and it is displayed. The user views this process automatically identification output of English translation by client. The LSTM handle long sequence of words which allows to consider sentence context and translate accurately to preserve meaning when the word order varies among languages. The LSTM handle difficult sentence and translate phrases which depended on previous words. The f1-score, precision, accuracy and recall are used for evaluating LSTM performance. The LSTM achieves 98.42% f1-score, 98.68% precision, 98.91% accuracy, 98.16% recall.