CAiRE in DialDoc21: Data Augmentation for Information-Seeking Dialogue System

TL;DR

This paper presents a data augmentation approach combined with training techniques using pre-trained language models to improve information-seeking dialogue systems, achieving promising results in the DialDoc21 competition.

Contribution

It introduces data augmentation methods and training strategies with pre-trained models specifically for dialogue systems, demonstrating their effectiveness in a competitive setting.

Findings

Achieved 74.95 F1 score in subtask 1

Achieved 60.74 Exact Match score in subtask 1

Achieved 37.72 SacreBLEU score in subtask 2

Abstract

Information-seeking dialogue systems, including knowledge identification and response generation, aim to respond to users with fluent, coherent, and informative responses based on users' needs, which. To tackle this challenge, we utilize data augmentation methods and several training techniques with the pre-trained language models to learn a general pattern of the task and thus achieve promising performance. In DialDoc21 competition, our system achieved 74.95 F1 score and 60.74 Exact Match score in subtask 1, and 37.72 SacreBLEU score in subtask 2. Empirical analysis is provided to explain the effectiveness of our approaches.