Multi-Task Learning for Domain-General Spoken Disfluency Detection in Dialogue Systems

TL;DR

This paper introduces a multi-task LSTM model for incremental, domain-general detection of spoken disfluencies in dialogue, improving accuracy and demonstrating strong generalization to synthetic datasets without retraining.

Contribution

The paper presents a novel multi-task LSTM approach for incremental disfluency detection that outperforms previous models and generalizes well across different datasets.

Findings

Outperforms prior neural approaches by ~10 percentage points on SWDA

Achieves good generalization to synthetic bAbI+ dataset without retraining

Supports real-time processing for downstream dialogue systems

Abstract

Spontaneous spoken dialogue is often disfluent, containing pauses, hesitations, self-corrections and false starts. Processing such phenomena is essential in understanding a speaker's intended meaning and controlling the flow of the conversation. Furthermore, this processing needs to be word-by-word incremental to allow further downstream processing to begin as early as possible in order to handle real spontaneous human conversational behaviour. In addition, from a developer's point of view, it is highly desirable to be able to develop systems which can be trained from `clean' examples while also able to generalise to the very diverse disfluent variations on the same data -- thereby enhancing both data-efficiency and robustness. In this paper, we present a multi-task LSTM-based model for incremental detection of disfluency structure, which can be hooked up to any component for incremental interpretation (e.g. an incremental semantic parser), or else simply used to `clean up' the current utterance as it is being produced. We train the system on the Switchboard Dialogue Acts (SWDA) corpus and present its accuracy on this dataset. Our model outperforms prior neural network-based incremental approaches by about 10 percentage points on SWDA while employing a simpler architecture. To test the model's generalisation potential, we evaluate the same model on the bAbI+ dataset, without any additional training. bAbI+ is a dataset of synthesised goal-oriented dialogues where we control the distribution of disfluencies and their types. This shows that our approach has good generalisation potential, and sheds more light on which types of disfluency might be amenable to domain-general processing.