Machine translation (MT) is the task of converting given text from one language to another language using automatic algorithms. Machine translation helps a lot to facilitate communication between people from different countries, e.g. in sharing knowledge or in traveling around. It is one of the most difficult problems in natural language processing, which requires an excellent bilingual ability and a comprehensive understanding of semantics and syntax. No wonder MT is categorized as AI-complete and drives the advances in sequence-to-sequence models.
This course will introduce basic concepts, past/present paradigms, and state-of-the-art advances of MT. This includes linguistic motivations, mathematical modeling, and deep learning tricks for MT. Assignments will provide an opportunity for building and managing a practical MT system.

- Probability and Statistics (MATH230)
- Artificial Intelligence (CSED342)
If you have taken other courses with similar contents as above, please consult me for recognition as prerequisites.
* Undergraduate students are welcome if they fulfill the prerequisites.
* This course is also suitable for exchange/foreign students.

Assignments: 60%
Midterm Exam: 30%
Quizzes: 10%

TBA

** No lecture on Tuesday, September 6th. The first lecture will be on Thursday, September 8th. **

1. Introduction & Primer
2. Word-based Models
3. Phrase-based Models
4. Sequence-based Models
5. Practical Pipeline
6. Decoding
7. Rule-based & Example-based
8. Midterm Exam
9. Collecting Parallel Data
10. Using Monolingual Data
11. Multilingual Models
12. Model Adaptation
13. Integrating Context
14. Human-in-the-Loop Setup
15. Speech Translation
16. Final Project Presentation

Assignments: work in a team of 2-3 people
Quizzes: answer a few questions about reading materials for each lecture

TBA