How to Predict a 5 Star Review

It may not be surprising to find terms such as "great" or "good" in 5-star reviews of a product. However, if a particular part of a product is repeatedly associated with negative ratings, it points to a problem that needs to be resolved.

Wordify, an online tool developed by the Data and Marketing Insights (DMI) Research Unit at Bocconi University, makes it easy to identify the words that delineate categories in textual data. It can be applied to any situation where you have a set of texts with labels associated with each of them and you want to know which terms best describe, or predict, each label. For instance, it helps discover the words most frequently used in a review associated with a particular rating.

This online tool can be used by anyone, not limited to data scientists. There are only a few steps to follow.

Simply go to the Wordify home page and upload an Excel sheet with two columns: one called "text", the other one called "label", with each row containing a text and its associated label. A short while later, you will receive the results of the Wordify analysis by email, telling the most indicative terms for each label.

The Data and Marketing Insights (DMI) Research Unit, headed by Scientific Director Dirk Hovy, combines modern marketing techniques with novel methods from machine learning, data science, and big data analysis. From social media analysis and better understanding of your target audience's demographics to machine-assisted decision making, affiliated researchers work on a broad range of topics.