An Orthonormal Basis for Topic Segmentation in Tutorial Dialogue

This paper explores the segmentation of tutorial dialogue into cohesive topics. A latent semantic space was created using conversations from human to human tutoring transcripts, allowing cohesion between utterances to be measured using vector similarity. Previous cohesionbased segmentation methods that focus on expository monologue are reapplied to these dialogues to create benchmarks for performance. A novel moving window technique using orthonormal bases of semantic vectors significantly outperforms these benchmarks on this dialogue segmentation task.