On a Guided Nonnegative Matrix Factorization

TL;DR

This paper introduces Guided NMF, a semi-supervised topic modeling approach that incorporates user-provided seed words to improve the quality of learned topics, demonstrating competitive results with minimal supervision.

Contribution

The paper proposes Guided NMF, a novel semi-supervised extension of NMF that uses seed words to guide topic discovery, addressing issues of redundant or less meaningful topics.

Findings

Guided NMF outperforms traditional unsupervised NMF in topic coherence.

The method requires minimal supervision to achieve competitive results.

Experimental results validate the effectiveness of seed word guidance in topic modeling.

Abstract

Fully unsupervised topic models have found fantastic success in document clustering and classification. However, these models often suffer from the tendency to learn less-than-meaningful or even redundant topics when the data is biased towards a set of features. For this reason, we propose an approach based upon the nonnegative matrix factorization (NMF) model, deemed \textit{Guided NMF}, that incorporates user-designed seed word supervision. Our experimental results demonstrate the promise of this model and illustrate that it is competitive with other methods of this ilk with only very little supervision information.