contribution.txt

Kevin Bowden, kkbowden@ucsc.edu
Johnnie Chang, cchan60@ucsc.edu
Robert Chen, hchen118@ucsc.edu
Bernardo Gonzalez, beaugonz@ucsc.edu

Project Work Distribution
We will now provide a summary of how the work was distributed for the project. We primarily
work as a team effort. Generally, all four of us met once a week, if not in person then remotely,
to discuss the project and how we intend to progress. As a team we decided on our proposal,
and all four of us worked equally on the project reports. Johnnie set up our initial repository,
and Robert found and added the data wrangling script we use in our project.
He and Bernardo also created the framework for our Naïve Bayes classifier and the preprocessing steps.
Johnnie and Kevin have had experience with other NLP work in the past, so they were responsible
for implementing the LIWC functionality, Bigrams, and Unigrams. Once we started to hit the core
of our code base we divided our tasks more; Robert added support for the scikit-learn logistic
regression model as well as the TF-IDF calculation, Bernardo added support for the scikit-learn SVM model,
Johnnie set up the driver functionality and the README file, Kevin built the function to get the most
discriminative features and verified documentation through the code. As a team we all performed a code
walk in which we cleaned up our code, performed unit testing, and optimized our system architecture.
Even though one of our members could not run the code, due to physical restrictions on their machine,
all members were present to see the output and perform group analysis.