Classification and Prediction
Barry deVille, Gurpreet Singh Bawa · 2021
This chapter considers repairs under warranty for a global truck manufacturer. A text analytics-based prediction and deployment application is developed for the truck manufacturer. This work was documented in two successive SAS User Conferences presentations. The chapter illustrates how ecommerce platforms can employ analytics to examine consumer reviews pertaining to the products bought from different partner sellers to understand the key reasons driving bad/good reviews. The features obtained from the preliminary analysis is considered. The chapter uses a binary target variable that reflects whether the review or feedback is a positive or negative sentiment. It adapts a standard text classification task with some modifications: text classification, ideally, is an exercise where the objective is to categorize document instances depending on the terms/tokens they are built from.
Cited by 9 works. doi:10.1002/9781119487142.ch6