Computer Aided Online Translation Teaching System

In the process of machine translation, similarity threshold is an important factor restricting retrieval and translation. When the fuzzy interval similarity threshold is low, the translation effect of the traditional online translation teaching system is poor, but if the threshold is too high, it will lead to retrieval difficulties and affect the translation progress. In view of this situation, a computer-aided online translation teaching system is designed. In the hardware design of the system, the content addressed memory is mainly studied to provide space for data storage and facilitate address search for the partition of translation cache package; In the software design, firstly, the neural loop network is established for deep learning, the natural language feature vector is extracted, and the source sentence is encoded to obtain the sentence vector with a certain length. After adding some algorithms and constraints, the decoding is completed with the help of neural network to eliminate the gradient imbalance in the process of translation ambiguity; Optimize the translation training model, reorder the extracted feature vectors, and optimize the translation results. In order to verify the effectiveness of the designed system, experiments are designed. The results show that when the fuzzy interval is less than 0.5, the performance of the designed system is significantly better than the traditional system. With the increase of similarity threshold, the performance gap between the two methods gradually narrows.