Java: Data Science Made Easy

Summary

The intent of this chapter was to illustrate how various data science tasks can be integrated into an application. We chose an application that processes tweets because it is a popular social medium and allows us to apply many of the techniques discussed in earlier chapters.

A simple console-based interface was used to avoid cluttering the discussion with specific but possibly irrelevant GUI details. The application prompted the user for a Twitter topic, a sub-topic, and the number of tweets to process. The analysis consisted of determining the sentiments of the tweets, with simple statistics regarding the positive or negative nature of the tweets.

The first step in the process was to build a sentiment model. We used LingPipe classes ...