Danial Soleimani - Senior student of artificial intelligence, Apadana Institute of Higher Education, Shiraz, Iran

This is a review on Intent Detection and Semantic Parsing for Navigation Dialogue Language Processing that shows some researches on voice-based human-machine interfaces in modern intelligent vehicles, particularly in navigation and infotainment applications. It highlights the role of Automatic Speech Recognition (ASR) in converting spoken audio to text, followed by the necessity of a Natural Language Processing (NLP) subsystem to comprehend contextual meaning and take appropriate actions. For the human-vehicle navigation dialogue application, two key tasks are identified: (۱) intent detection, determining if a sentence is navigation-related, and (۲) semantic parsing, extracting crucial information like point-of-interest destinations. The study proposes a Recurrent Neural Network (RNN) architecture, exploring considerations such as joint vs. separate models, context window vs. sequence-to-sequence translation, and alternative model hyperparameter selections. Experiments using benchmark datasets show promising results, with the proposed solution achieving high accuracies of ۹۸.۲۴% for intent detection and ۹۹.۶۰% for semantic parsing, outperforming state-of-the-art methods, particularly on the CU-Move dataset

RNN, CNN, NLP, ASR, SNR, Deep Neural Network, LSTM