twitterRStorm

This repository contains code used to prototype a streaming framework for analysing Tweets about Comcast using the RStorm packages.

This is primarily being created for a presentation/tutorial at Interface 2015.

Purpose

The purpose of the stream is to simulate/prototype a streaming framework for a someone wishing to analyze tweets matching a given topic, in this case Comcast. The stream will:

1. Keep track and visualize (with a wordcloud) common terms associated with the topic.
2. Classify and visualize the polarity (positive/negative/neutral) of tweets and visual common words in each class.
3. Keep track of the proportion of positive tweets as time goes on.
4. Visualize and report the rate of tweets in a given time frame.

Dependencies

This project requires several R packages:

• RStorm
• twitteR
• sentiment*
• wordcloud
• dplyr
• tidyr

Most of these packages can be installed from CRAN, but sentiment and one of its dependencies need to be installed from source:

install.packages("tm")
install.packages("http://cran.r-project.org/src/contrib/Archive/Rstem/Rstem_0.4-1.tar.gz",
	             repo = NULL, type = "source")
install.packages("http://cran.r-project.org/src/contrib/Archive/sentiment/sentiment_0.2.tar.gz",
	             repo = NULL, type = "source")

Status

Current Contents:

File	Description
comcast_dash	Folder containing shiny demo
presentation	Folder containing presentation source code
tutorial	Folder containing tutorial source code
stormr	folder containing Storm package example
twitteRStorm.R	Standalone R Script of tutorial.
README.md	Readme file for project.
license.md	License document for project.

Viewing the Dashboard

The dashboard is designed to prototype how a dashboard for monitoring tweets might took. You can run the app locally by calling shiny::runApp() in R from within the comcast_dash directory.

Alternatively, you can run the app from shinyapps.io using the link: https://raffled.shinyapps.io/comcast_dash.