An Approach to Natural Language Intention Understanding of Civil Aviation Passengers Based on DIET Architecture

With the development of science and technology, more and more products appeared in the airport terminal to provide services for the passenger in civil aviation. However, due to the lack of understanding of natural language by scientific and technological equipment, the repeated consultation of the same question often fails to give satisfactory answers to the passengers, which reduces the service efficiency and friendliness. In order to provide better services for passengers in civil aviation, we aim at the common problems of passengers in the airport environment, simulate the possible service demands of passengers, analyze the key feature data in the demands, and combine the joint identification model of DIET to realize the accurate understanding of the intention in the natural language of passengers. In this paper, Python tools are used for making simulation experiments of natural language and intention understanding to verify the feasibility of the implementation method in this paper.