A comparison of two state-of-the-art dynamic topic models that solve the consumer complaint classification exercise.
Customer Feedback about products and services provide valuable information about customer satisfaction. They provide information on what needs to be improved throughout product development. Dynamic thematic models in business intelligence can identify key product qualities and other satisfaction factors, group them into categories, and evaluate how business decisions translated into customer satisfaction over time. This is very valuable information not only for product managers.
This article will compare two of the latest topic models for classifying customer complaint data. BERTopic by Maarten Grootendorst (2022) and the recent FASTopic by Xiaobao Wu et al (2024) presented last year NeurIPSare the current leading models for thematic analysis of customer data. For these models, we will explore in Python code:
- how to do it effectively preprocessing data
- how to train a Bigram Theme Model for the analysis of customer complaints
- how to model thematic activity over time.
