Design of a Cabin English Automatic Broadcasting Detection System Based on Deep Learning

Cabin English is a commercial English designed explicitly for overseas passengers to help them better understand platform information and stay abreast of current situations. As Cabin English can guide foreign tourists to make the right decisions to make the expression in Cabin English automatic broadcasting more accurate and clear, we propose a Cabin English broadcasting grammar correction detection system based on deep learning. We established a Cabin English broadcasting grammar correction model using deep belief network technology and applied it to the automatic broadcasting grammar detection and correction system. After testing, the frame loss rate of this model is as low as 0.11%, and the system can also be continuously updated to ensure the accuracy and reliability of voice data. The above results indicate that the deep belief network can perform grammar corrections for Cabin English and improve the accuracy of Cabin English broadcasting through feedback. This has significant implications for the application and development of commercial English in our country.