CSE 40657/60657
Project

Throughout the course, you will develop an application of NLP techniques to a topic of interest to you.

• You may work alone or in pairs as long as you clearly define which person did which parts of the work. Each person should contribute equally. (If an undergraduate and graduate student work together, the graduate student should contribute roughly 2/3.)
• It’s not acceptable to do the same work for this project and another class’s project, but it’s acceptable (and encouraged) for this project to relate to another project as long as the boundary is clearly defined.

PM1: Idea

For the first milestone, you will choose a problem to work on (and collaborator, if any). The deadline is intentionally set too early for you to know how to solve your problem. We just want to know what you're interested in! It is okay to change later on. Length guideline: ½–1 page per undergraduate, 1–2 pages per graduate. Graduate students should format the proposal like the introduction or motivation section of a conference paper. Grade: 30 points. Reasonable proposals get full credit, and the grader may allow you to fix problems and resubmit.

Your proposal should include the following points:

• What are you trying to do? Describe your goal without using any jargon.
• Write (or draw) an example of the envisioned input/output or user interaction.
• If your project is successful, how will it be helpful to people?

Please submit your idea as a PDF file in Canvas under "PM1".

PM2: Data and Baseline

For the second milestone, you should have everything ready that you need to start experimenting with solutions to your problem. Describe what you've done so far. Length guideline: ½–1 page per undergraduate, 1–2 pages per graduate. Graduate students should format their description like the data/baseline sections of a conference paper. Grade: 30 points.

For 95% of projects, this means that you must answer the following:

• What data will you use? Include URLs or LDC catalog numbers. Describe whatever preprocessing steps you used on the data (e.g., cleaning, tokenization).
• What metric(s) will you use to measure success?
• What baseline method will you compare against? This should be something that you can implement in about one hour.
• How well does your baseline method perform (using your evaluation metric)?

For a minority of projects, the above will not apply; please speak to Prof. Chiang about what to submit.

Please submit your work as a PDF file in Canvas under "PM2".

PM3: Presentation

This year you have a choice between giving a talk in class or presenting a poster. In either case, you should present:

• Goal: What are you trying to do (using no jargon)?
• Method: How are you trying to do it (briefly)?
• Experiments: How well does it work (so far)?
• Conclusions: What did you learn from your experiments (so far)?

In-class presentation: Your presentation should be no more than 5 minutes per person. It is okay if some of the work presented isn’t finished yet. Presentation dates are scattered throughout the second half of the semester. If you choose an early date, you will get feedback that you can actually implement! Grade: 30 points, based on presentation quality rather than research results. After your presentation, please submit your slides in Canvas under "PM3".

Poster presentation: The poster session will be on 12/11 from 4:15pm to 6:15pm. Single-person teams will present during either the first or second hour; two-person teams will present during both hours. If you want us to print your poster for you, please submit your poster as a 24"×36" PDF file in Canvas under "PM3" by 12/5. Otherwise, you will be responsible for printing your own poster. (Printing on 12 or 16 sheets of letter paper is acceptable.) We will provide poster boards, pushpins, and easels for mounting your posters. After your presentation, please submit your poster in Canvas under "PM3".

PM4: Report

Length guideline: 2–4 pages per undergraduate student, 4–8 pages per graduate student. Graduate students should format the report like a conference paper (e.g., ACL style). Grade: 60 points.

• Goal: What are you trying to do (using no jargon)? Give an example of inputs/outputs or user interaction. How might it be helpful to people?
• Method: How are you trying to solve the problem? What are existing approaches to the problem?
• Experiments: What data do you use? What metric(s) do you use to measure success? What baseline method do you compare against? How well do your methods perform compared with the baseline, and why?
• Conclusions: What did you learn from your experiments?
• Replicability: Submit (or include links to) all the code that you wrote and all the data that you used.

Please submit your report as a PDF file in Canvas under "PM4".