Skip to content
Jason Baldridge edited this page May 6, 2013 · 1 revision

Course project, Final report

DUE: May 13, 2013, 1pm CST

Preamble

This is your final report! The main emphasis should be on tightening up any coding and results, providing documentation in code and on your repository's wiki, and writing up the work formally in your paper submission.

Don't hesitate to ask the instructor if you would like further clarification on any of the requirements!

Submitting your solutions

Write up what you did in as a file <lastname>_<firstname>_final_report.pdf. Submit this on Blackboard.

Note: All final reports will be posted to the course website. If you would prefer your report not be posted, please let the instructor know.

Your code submission will be contained in your project code repository. You should tag the submission version of your repository as "ANLP-Course-Final-Report".

Requirements

  • Content: Write your report as if it were being submitted to the ACL demos session, not just in formatting, but in form and content.
    • As an example of a nice report to follow, see Liu and Baldwin (2012), langid.py: An Off-the-shelf Language Identification Tool
    • Make sure to incorporate feedback from project phase five.
    • Evaluation is important, so don't skimp on it. It is far better to have fewer capabilities and meaningful evaluation than to have all the whiz bang stuff but no clear evidence that it works as intended.
  • Length and formatting. Your report must be written using the ACL System Demonstrations style files (you are free to use LaTeX or Word). The maximum length is 6 pages, not including references (you may have an extra page for those). (Note: a 3-4 page submission is unlikely to get a good grade.) (Your final submission will be in the same style, but conforming to the full submission instructions described at the above link.)
  • References. You should have at least ten (more is fine), at least three of which do not come from UT Austin. These should be academic papers of relevance to what you have done. If you are citing blogs, manuals, news articles, etc, those should just be footnotes and don't count as references. If you are having trouble finding related work, just let me know ahead of time and I'll help you track things down or discuss what you should be looking for.
  • Documentation in code. All of the major methods, classes, traits, etc. should be documented using Scaladoc style. Minimally, they should describe the relevant piece of code, and it would be nice if the inputs and outputs are also described using flags like @param. (Hopefully, you've mostly done this as part of phase five.)

Rubric

Your submission will be scored using the following rubric. Qualities of full-point submissions are given below each area.

It is fine to use text from project phase five for this submission -- as before, you are basically iteratively refining that document.

  • Structure & Writing: 30
    • Formatting & Length: 5
    • Spelling/Grammar: 5
    • Use of figures and tables: 5
    • Clarity of exposition: 10
    • References (minimum 10, at least 7 from non-UT Austin authors): 5
  • Content: 30
    • Creativity: 5
    • Methodology: 5
    • Evaluation rigor: 10
    • Analysis/Interpretation of findings: 10
  • Code: 20
    • In-code documentation: 5
    • Supporting documentation (READMEs, wiki how-tos): 5
    • Complexity of implementation (both in terms of NLP/ML methods and programming complexity): 10
  • Overall quality: 20

Please look at the write-ups by the others for project phase five to get ideas for how you should structure and present your own.

Clone this wiki locally