Entry Date:
October 25, 2017

Arabic Speech and Language Processing

Principal Investigator Regina Barzilay

Co-investigators James Glass , Tommi Jaakkola


This project aims to develop speech and language processing technologies that will support natural interaction via spoken language. Specific research objectives include developing technologies that can address low-resource issues related to Arabic and language processing methods to process user-generated content, understanding and dialogue capability to enable sophisticated spoken language interfaces, and data-driven speech and language methods that can learn through interaction and feedback. A practical goal for this project is the application of these technologies, e.g., to support Al-Jazeera video processing, for spoken language support for a Doha-specific tourist guide, and for a system providing information for the 2022 World Cup in Doha.