Topic subject creation using unsupervised learning for topic modeling

TL;DR

This paper compares NMF and LDA algorithms for topic mining and labeling in retail customer communications, proposing automated methods for assigning topic labels to better understand customer inquiries.

Contribution

It introduces a comparative analysis of NMF and LDA for topic extraction and develops automated labeling techniques for improved topic characterization.

Findings

NMF and LDA show different strengths in topic mining performance.

Automated labeling methods effectively assign subject labels to topics.

The approach enhances understanding of customer inquiries in retail.

Abstract

We describe the use of Non-Negative Matrix Factorization (NMF) and Latent Dirichlet Allocation (LDA) algorithms to perform topic mining and labelling applied to retail customer communications in attempt to characterize the subject of customers inquiries. In this paper we compare both algorithms in the topic mining performance and propose methods to assign topic subject labels in an automated way.