Cold-Start Active Correlation Clustering
This work addresses a specific challenge in active learning for clustering, but it appears incremental as it builds on existing correlation clustering methods.
The paper tackled the problem of active correlation clustering in a cold-start scenario where no initial pairwise similarities are available, proposing a coverage-aware method to encourage diversity early in the process, and demonstrated its effectiveness through synthetic and real-world experiments.
We study active correlation clustering where pairwise similarities are not provided upfront and must be queried in a cost-efficient manner through active learning. Specifically, we focus on the cold-start scenario, where no true initial pairwise similarities are available for active learning. To address this challenge, we propose a coverage-aware method that encourages diversity early in the process. We demonstrate the effectiveness of our approach through several synthetic and real-world experiments.