Airport mobile information access with spoken query over wireless networks

The today's mobile communication networks and modern end-user devices provide an access to the Internet while on the move. The paper builds a model of airport mobile information service accessibility with spoken query. A system is developed and tested through the prototype that applies state-of-the-art Speech Recognition with robust feature extraction and knowledge-based Information Retrieval processing for spoken query. The model adopts a traditional Client-Server architecture. For the Server part, a speak recognizer system is implemented on the basis of ETSI standard DSR front-end with cepstral mean subtraction (CMS) method combined with CW and cepstral derivatives; and a knowledge-based Information extraction solution is developed for fast retrieval of the most relevant documents. For the Client part, a Graphical User Interface (GUI) is designed to show the focused part of the query result. The two parts are integrated into a full spoken query system. The system has the flexibility for being ported to any airport.