New Data: Revelio Labs Sentiment Scores

An additional dataset has recently been added to the U of T's Revelio Labs data license. 

New: Sentiment Scores

This dataset contains employee sentiment scores that were generated using a sentiment model, which uses Natural Language Processing to capture employee sentiment on specific topics such as management and diversity. For each review, a weighted sentiment score is computed, assigning a positive (negative) score to topics that had an overall positive (negative) impact on the review. These scores are then aggregated to arrive at a company-wide sentiment score.

Accessing the data

The Map & Data Library has created a detailed tutorial on the process of applying for and using the data which is hosted in SciNet's supercomputing environment.