KPT: Keyword-guided Pre-training for Grounded Dialog Generation

TL;DR

KPT is a self-supervised pre-training method that enhances grounded dialog generation by extracting keywords from dialogs, enabling models to better incorporate various types of external knowledge without additional annotations.

Contribution

This work introduces KPT, a novel keyword-guided pre-training approach that leverages dialog-internal keywords for improved knowledge-grounded response generation across multiple knowledge sources.

Findings

KPT outperforms state-of-the-art methods on multiple knowledge-grounded tasks.

It effectively handles diverse knowledge types like graphs, personas, and passages.

The method works well with large-scale, unlabeled dialog data.

Abstract

Incorporating external knowledge into the response generation process is essential to building more helpful and reliable dialog agents. However, collecting knowledge-grounded conversations is often costly, calling for a better pre-trained model for grounded dialog generation that generalizes well w.r.t. different types of knowledge. In this work, we propose KPT (Keyword-guided Pre-Training), a novel self-supervised pre-training method for grounded dialog generation without relying on extra knowledge annotation. Specifically, we use a pre-trained language model to extract the most uncertain tokens in the dialog as keywords. With these keywords, we construct two kinds of knowledge and pre-train a knowledge-grounded response generation model, aiming at handling two different scenarios: (1) the knowledge should be faithfully grounded; (2) it can be selectively used. For the former, the grounding knowledge consists of keywords extracted from the response. For the latter, the grounding knowledge is additionally augmented with keywords extracted from other utterances in the same dialog. Since the knowledge is extracted from the dialog itself, KPT can be easily performed on a large volume and variety of dialogue data. We considered three data sources (open-domain, task-oriented, conversational QA) with a total of 2.5M dialogues. We conduct extensive experiments on various few-shot knowledge-grounded generation tasks, including grounding on dialog acts, knowledge graphs, persona descriptions, and Wikipedia passages. Our comprehensive experiments and analyses demonstrate that KPT consistently outperforms state-of-the-art methods on these tasks with diverse grounding knowledge.