COMP 396 Undergraduate Research Project Proposal Form: "Applying French Morphological Rules...". (Prof. R. Rose)

INSTRUCTIONS - PROFESSORS: Fill out Sections A & B then submit this form online. (You will receive an email copy of the form. The Office for Undergraduate Research in Science will also post the project online, indicating whether the project is open for students to apply or taken.) DONE
INSTRUCTIONS - STUDENTS: You may receive this form by email, or you may download it after it has been posted online. Either way, print this form. Complete and sign Section C on the hardcopy. Ask your supervisor to sign Section D. Take it to the department corresponding to the course number in Section A (this may or may not be your own department). Do not register for a '396' course on Minerva until you receive departmental permission.
INSTRUCTIONS - DEPARTMENTS: After the unit chair/director/designate approves (or not) this project, notify student. If approved, please give student permission to register on Minerva, and fax this form (with signatures) to the Office for Undergraduate Research in Science at 514-398-8102.
QUESTIONS OR FEEDBACK? Contact Victor Chisholm by email, or phone 514-398-5964.
Name: Dr. Richard Rose
Email: rose [at]
Supervisor's department: None of the above
Supervisor's department (if none of the above) Electrical and Computer Engineering
Course number: COMP 396 (Computer Science)
Term: Fall 2009-2010
Project start date: September 15, 2009
Project end date: December 1, 2009
Project title: Applying French Morphological Rules to Improve French-English Statistical Machine Translation Systems
Project description: A current topic of interest in statistical machine translation (SMT) is in combining purely statistic knowledge derived from bilingual text corpora with rule based linguistic knowledge. It is thought that there may be many instances where morphological variation in French is not well modeled by existing French-English statistical machine translation (SMT) systems. This project will investigate this issue by applying a set of French language morphological rules to augment lists of phrases that are generally acquired automatically from data in SMT training. The project consists in (1) choosing a/some performance measure(s) for "goodness of translation", (2) choosing a framework (an already implemented purely statistical automatic translater), (3) choosing a set of morphological rules to apply, (4) implementing the morphological rules in the framework, and finally (5) comparing the results of the original framework and the modified framework using the chosen performance measure(s). We expect the overall results of the modified framework to yield a better translation than the original framework. Due to the widespread use of statistical machine translation, the results could later be applied on other frameworks.
Prerequisite: 1 term completed at McGill + CGPA of 3.0 or higher; or permission of instructor.
Other prerequisite, if applicable:
Grading scheme (The final report must be worth at least 50% of final grade): 75% final report (minimum 10 pages) and 25% presentation.
Other project information:
Project status - This project is: Taken. The professor has no more '396' projects this term.
How students can apply: N/A; this project is filled.
If other, please specify:  
Ethics, safety, and training
Which of the following, if any, is involved? One or more of the following
Animal subjects [ ]
Human subjects [ ]
Biohazardous substances [ ]
Radioactive materials [ ]
Handling chemicals [ ]
Using lasers [ ]
Supervisors are responsible for the ethics and safety compliance of undergraduate students.
Do not complete this section unless/until the student is identified.
McGill ID:
Email (first.last [at]
Program (E.g., B.Sc. Maj. Chem. Min. Biol.):
Level (U0 / U1 / U2 / U3):
Student signature - I have not applied for another '396' course in this term:  
Do not complete this section unless/until the student is identified.
Supervisor: I give my permission for the student identified in section C to register for this project under my supervision.
Supervisor's signature:  
Unit chair/director/designate: I certify that this project conforms to departmental requirements for 396 courses.
Unit chair/director/designate's name:  
Unit chair/director/designate's signature: