1. Amazon vs. Google
2. Remember the workflow?
It's time to put your newly acquired text mining skills to the test! Remember our awesome text mining workflow from chapter 1? This is the basic mental map of defining a problem, then transitioning from an unorganized state to an organized state. In doing so, you hopefully can reach a conclusion or novel insight.
3. A case study in HR analytics
Chapter 4 is a case study where you pretend you work in HR Analytics and have to help your recruiters understand how external candidates may perceive your organization compared to a competitor.
4. A case study in HR analytics
So you get to apply this workflow to real Amazon and Google employee reviews.
I manually gathered 2000 anonymous online reviews from current and former employees from both companies. Performing a sound text mining project usually requires following 6 steps in the context of this workflow.
The first two steps include defining the goal or problem you want to work.
5. A case study in HR analytics
Your problem statement should be specific. "Does Amazon or Google have better work-life balance and pay perception online?" In step two, you identify the text to be used. I did the heavy lifting of reading 2000 reviews
6. A case study in HR analytics
and putting them into positive or negative buckets for a total of four corpora. A positive and negative group for Amazon and another set of positive and negative for Google.
7. Let's practice!