Deep Descriptive Clustering

TL;DR

This paper introduces deep descriptive clustering, a method that combines deep learning for complex data with interpretable explanations, achieving high-quality clusters and explanations simultaneously.

Contribution

It proposes a novel framework that learns sub-symbolic representations and generates symbolic explanations, improving clustering interpretability for complex data types.

Findings

Outperforms baselines in clustering accuracy

Provides high-quality, orthogonal cluster explanations

Enhances clustering with explanation-informed constraints

Abstract

Recent work on explainable clustering allows describing clusters when the features are interpretable. However, much modern machine learning focuses on complex data such as images, text, and graphs where deep learning is used but the raw features of data are not interpretable. This paper explores a novel setting for performing clustering on complex data while simultaneously generating explanations using interpretable tags. We propose deep descriptive clustering that performs sub-symbolic representation learning on complex data while generating explanations based on symbolic data. We form good clusters by maximizing the mutual information between empirical distribution on the inputs and the induced clustering labels for clustering objectives. We generate explanations by solving an integer linear programming that generates concise and orthogonal descriptions for each cluster. Finally, we allow the explanation to inform better clustering by proposing a novel pairwise loss with self-generated constraints to maximize the clustering and explanation module's consistency. Experimental results on public data demonstrate that our model outperforms competitive baselines in clustering performance while offering high-quality cluster-level explanations.