davidprichard / sentimental-stats

Sentimental Stats is designed to collect useful statistics from public opinions of a sports team's performance.

Sentimental Stats

This was originally a term project for ITCS 5102 "Survey of Programming Languages" at UNCC, completed mostly to learn Scala. The original project was very limited, collecting stats from sportsdatabase.com, collecting Tweets referencing the a particular team, estimating the average sentiment over time, and plotting those stats.

For now, I'll focus on improving the sentimental data by using more reliable sources of data than Twitter and sportsdatabase.com, and building up a broader slice of metrics like margin-of-victory that are more useful for predicting performance. I hope to graph the results with something better than JFreeChart/ScalaChart as well.

How to Run

1. Download the project or clone the repo with git clone https://github.com/4102/Sentimental-Stats.git

2. Download and install sbt

3. To compile and run, navigate to the repo's root directory from the shell of your choice and enter sbt run ...* To run with command-line arguments, wrap them in quotes like so: sbt "run arg1 arg2

How to View/Edit

1. Download and install IntelliJ.
2. Select File > Open... and navigate to the project if it doesn't detect it on start.
3. Install the Scala plugin when prompted, or do it manually.
4. Add the lib/ directory to your dependencies/classpath:

• Go to File > Project Structure > Modules and select the Dependencies tab
• Select the + symbol, choose jars or directories, and finally navigate to lib/. Phew.
  • You have to follow the same process to add the individual jars in the root of lib, as IntelliJ can't identify them on its own. This is frustrating but necessary.

5. The project doesn't reliably compile or run in IntelliJ, so I strongly recommend using sbt for that instead.

Dependencies

Sentimental Stats uses:

• Standford's NLP library
• jsoup