SGD-X: A Benchmark for Robust Generalization in Schema-Guided Dialogue Systems

TL;DR

SGD-X is a new benchmark that tests the robustness of schema-guided dialogue systems to linguistic variations, revealing current models' limitations and proposing a data augmentation method to enhance generalization.

Contribution

The paper introduces SGD-X, a benchmark for evaluating robustness to schema variations, and proposes a data augmentation technique to improve model generalization.

Findings

State-of-the-art models struggle with schema variations.

Schema sensitivity impacts joint goal accuracy.

Data augmentation improves robustness.

Abstract

Zero/few-shot transfer to unseen services is a critical challenge in task-oriented dialogue research. The Schema-Guided Dialogue (SGD) dataset introduced a paradigm for enabling models to support any service in zero-shot through schemas, which describe service APIs to models in natural language. We explore the robustness of dialogue systems to linguistic variations in schemas by designing SGD-X - a benchmark extending SGD with semantically similar yet stylistically diverse variants for every schema. We observe that two top state tracking models fail to generalize well across schema variants, measured by joint goal accuracy and a novel metric for measuring schema sensitivity. Additionally, we present a simple model-agnostic data augmentation method to improve schema robustness.