BP-Seg: A graphical model approach to unsupervised and non-contiguous text segmentation using belief propagation
This addresses the problem of semantic text segmentation for downstream applications, but it appears incremental as it builds on existing graphical model techniques.
The paper tackled unsupervised text segmentation by proposing BP-Seg, a graphical model approach using belief propagation to group both adjacent and distant sentences based on semantic similarity, and it demonstrated favorable performance compared to competing methods on long-form documents.
Text segmentation based on the semantic meaning of sentences is a fundamental task with broad utility in many downstream applications. In this paper, we propose a graphical model-based unsupervised learning approach, named BP-Seg for efficient text segmentation. Our method not only considers local coherence, capturing the intuition that adjacent sentences are often more related, but also effectively groups sentences that are distant in the text yet semantically similar. This is achieved through belief propagation on the carefully constructed graphical models. Experimental results on both an illustrative example and a dataset with long-form documents demonstrate that our method performs favorably compared to competing approaches.