DFlow: Diverse Dialogue Flow Simulation with Large Language Models

TL;DR

This paper introduces DFlow, a novel data simulation approach that enhances task logic diversity in dialogue datasets using large language models to generate decision tree-structured task plans, improving dialogue agent training.

Contribution

The paper presents a new method for generating diverse dialogue flows based on task execution logic, addressing a gap in existing data simulation techniques.

Findings

Generated 3,886 dialogue flows across 15 domains

Models trained on this data outperform baselines including GPT-4

Enhanced task logic diversity improves next action prediction accuracy

Abstract

Developing language model-based dialogue agents requires effective data to train models that can follow specific task logic. However, most existing data simulation methods focus on increasing diversity in language, topics, or dialogue acts at the utterance level, largely neglecting a critical aspect of task logic diversity at the dialogue level. This paper proposes a novel data simulation method designed to enhance the diversity of synthetic dialogues by focusing on task execution logic. Our method uses LLMs to generate decision tree-structured task plans, which enables the derivation of diverse dialogue trajectories for a given task. Each trajectory, referred to as a "dialog flow", guides the generation of a multi-turn dialogue that follows a unique trajectory. We apply this method to generate a task-oriented dialogue dataset comprising 3,886 dialogue flows across 15 different domains. We validate the effectiveness of this dataset using the next action prediction task, where models fine-tuned on our dataset outperform strong baselines, including GPT-4. Upon acceptance of this paper, we plan to release the code and data publicly.