Schedule    Home    Syllabus   

All of these deadlines must be met to receive credit on the course project:

  • By February 7: meet with me to discuss your project, before writing your proposal
  • February 19: Submit a written project proposal. Information about the proposal is *here*
  • March 21: Submit a written progress report. Information about the progress report is *here*
  • April 16: Submit a draft of your report. Follow the guidelines and use the style files from ACL 2012, available *here* The body of your paper should be 4 pages, excluding references. You may have any number of pages of references. In addition, you should include a one-page appendix of interesting examples.
  • April 23: Submit your final report.
  • April 23-25: Project presentations.
Your project should be non-trivial and interesting, yet feasible given the time frame.

Projects should be done in groups of 2 or 3.

The project may be related to your own research, but should represent additional work (see the proposal guidelines).

There are four options for the project:

* A corpus annotation project: This will involve developing annotation instructions, gathering a corpus, performing a training round of annotation, discussing the results with each other, revising the annotation instructions, and then annotating a fresh test set. Inter-annotator agreement should be reported. The amount of annotated data does not need to be large. We will have a class on annotation, including measuring inter-annotator agreement.
* Existing data and task: Implement and evaluate an algorithm using publicly available data, where you define the task as in previous work, and you attempt to improve upon previous results. Some pointers to available data will be placed on the schedule. You may also search for others.
* Existing data, new task: Implement and evaluate an algorithm using publicly available data, but use the data in a novel way.
* Application system: Use opinion/sentiment knowledge to improve an application system such as a summarization or information extraction system. Processing may be fully automatic, or your system may take manual annotations as input.