Keyword and named entity recognition on air traffic control text

• Main Author: Tay, Nikole Qiwei
• Other Authors: Chng Eng Siong
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence; Engineering::Computer science and engineering::Computing methodologies::Document and text processing
• Online Access: https://hdl.handle.net/10356/144503
• Institution: Nanyang Technological University

Aviation safety is utmost critical to ensure that passengers and flight crew are safe while in flight. Safety always comes first as lives are involved on every flight. It has always been the most important aspect of air travel. Past incidents of miscommunication between pilots and air traffic controllers had resulted in ill-fated accidents that could have been prevented. It is therefore crucial that we uncover solutions to prevent mishaps from occurring in the skies and at airports. This report will discuss the application of Named Entity Recognition (NER), a Natural Language Processing (NLP) method on Air Traffic Control (ATC) conversations to improve and attain higher precision in the communication between pilots and air traffic controllers. An ATC dataset will be generated based on grammar rules through the usage of OpenFST and the Thrax Compiler. Thereafter, this ATC dataset will be used to train a model using a Bi-LSTM-CRF based NER system. A demo website which allows the input of ATC text features the working application of NER by returning the highlighted and annotated name entities.