# Hierarchical RNN with Static Sentence-Level Attention for Text-Based Speaker Change Detection

## TL;DR

This paper introduces a hierarchical RNN with static sentence-level attention for text-based speaker change detection, demonstrating improved performance over previous methods in dialog transcript analysis.

## Contribution

The paper proposes a novel hierarchical RNN model with static sentence-level attention specifically designed for text-based speaker change detection, addressing a gap in existing audio-focused research.

## Key findings

- Neural network models outperform feature-based approaches.
- Attention mechanism significantly improves detection accuracy.
- Model effectively handles dialog transcripts without speaker identities.

## Abstract

Speaker change detection (SCD) is an important task in dialog modeling. Our paper addresses the problem of text-based SCD, which differs from existing audio-based studies and is useful in various scenarios, for example, processing dialog transcripts where speaker identities are missing (e.g., OpenSubtitle), and enhancing audio SCD with textual information. We formulate text-based SCD as a matching problem of utterances before and after a certain decision point; we propose a hierarchical recurrent neural network (RNN) with static sentence-level attention. Experimental results show that neural networks consistently achieve better performance than feature-based approaches, and that our attention-based model significantly outperforms non-attention neural networks.