COBRA: A Fast and Simple Method for Active Clustering with Pairwise Constraints

TL;DR

COBRA is a fast, active clustering method that over-clusters data with K-means and then merges clusters using pairwise constraints, minimizing queries and outperforming existing methods in quality and speed.

Contribution

Introduces COBRA, a novel active clustering algorithm that efficiently merges over-clusters with minimal pairwise constraints, leveraging transitivity and entailment.

Findings

Outperforms state-of-the-art in clustering quality

Faster runtime compared to existing methods

Does not require pre-specifying number of clusters

Abstract

Clustering is inherently ill-posed: there often exist multiple valid clusterings of a single dataset, and without any additional information a clustering system has no way of knowing which clustering it should produce. This motivates the use of constraints in clustering, as they allow users to communicate their interests to the clustering system. Active constraint-based clustering algorithms select the most useful constraints to query, aiming to produce a good clustering using as few constraints as possible. We propose COBRA, an active method that first over-clusters the data by running K-means with a $K$ that is intended to be too large, and subsequently merges the resulting small clusters into larger ones based on pairwise constraints. In its merging step, COBRA is able to keep the number of pairwise queries low by maximally exploiting constraint transitivity and entailment. We experimentally show that COBRA outperforms the state of the art in terms of clustering quality and runtime, without requiring the number of clusters in advance.