Initial Results for Final Paper Assignment

Due Date: Monday, November 19, 2018

Submission Instructions: Submit digitally (on Box) and be prepared to discuss in our next class

Assignment Description: For this assignment, you will turn in a file (in your folder) that represents the results or output of your initial computation. No interpretation of the results is required, but you should update your README file with details about the computational method you are using, and any preferences/settings you have decided on that may have affected your result. The specific structure of your results files will vary depending on your computational method but, in general, a spreadsheet (csv, excel, etc.) with rows and columns for your data is the preferred format. Here are some examples:

Collocations on a corpusA list of term pairs and their collocations scoresUpdate readme with specific settings such as which collocations measure you used
Topic modelsSpreadsheet of each topic model and the model's top terms; spreadsheet of each text and a score for how well it fits each modelMALLET can export these data as csv files
TF-IDFA folder of spreadsheets, one for each document, listing each term in the document and its TF-IDF scoreUpdate readme with specific settings such as ignored words, normalization, etc.
Deriving a count or percentile scoreOne row per text, one column per measureIdeal for something like sentiment scores

When in doubt, your results should make it relatively easy to match your numerical data to a particular text. If you have any questions about a preferred results format for your computation, please email me directly.

No hard copy of these results is required, but you should come to our next class prepared to discuss your project, as we will do a quick, informal check-in. The main purpose of this assignment is to make sure you have decided on an initial computation, and that you have set up your project in such a way (data and code/software) by now that your are able to produce an initial computation. You may find that change some aspect of the computation for your final paper (code, data, settings, etc.) but this version will represent a dry run so that you can address any unforeseen issues that may arise.