Our New R Package for Tapping into IBM Watson’s Sentiment Analysis API

IBM_Vintage_DataCenter.png

Introducing IBMsentR, a simple R package for textual sentiment and tone analysis through IBM Watson

In the spirit of our other R packages like FBinsightsR, we’ve released IBMsentR. The goal of these packages is simple: make slightly laborious API interactions a lot easier.

We’ve written extensively about how to use R & different cloud based sentiment analysis services, what they cost, and even how, side by side, they can be used to achieve some pretty impressive results. While doing that discovery and analysis we wrote a few functions to tap into the IBM Watson sentiment analysis services. We put these into a package to do two things:

  1. Ibm_sent - analyse sentiment of a given chunk of text or string

  2. Ibm_tone - analyse the tone of a given chunk of text or string

The IBMsentR package saves time because it returns a tibble (data frame) every time. No messing around neither with JSON nor lists.

 Sample text string for analysis

Sample text string for analysis

 Sample result from IBM Watson API

Sample result from IBM Watson API

In version 1 of IBMsentR, you’ll need your own IBM account credentials and you’ll only be able to submit one text string per function call. But we’ll be expanding its functionality soon to allow for larger submissions.

Give it a try, we hope you find it useful.

IBMsentR on GitHub