Warm-up – data exploration

Let's get things moving with a tiny example. Let's look at this tiny reviews corpus:

text <- c("The food is typical Czech, and the beer is good. The service is quick, if short and blunt, and the waiting on staff could do with a bit of customer service training",
          "The food was okay. Really not bad, but we had better",
          "A venue full of locals. No nonsense, no gimmicks. Only went for drinks which were good and cheap. People friendly enough.",
          "Great food, lovely staff, very reasonable prices considering the location!")

We will do some simple analysis here, which will help us appreciate some of the subtleties of sentiment analysis.

Working with tidy text

For this, we will use the tidytext package. This package is built on the philosophy of tidy data, introduced by Hadley Wickham in his 2014 paper (https://www.jstatsoft.org/article/view/v059i10). A dataset is tidy if the following three conditions are satisfied:

• Each variable is a column
• Each...