Beyond Domain APIs: Task-oriented Conversational Modeling with Unstructured Knowledge Access

TL;DR

This paper enhances task-oriented dialogue systems by integrating unstructured external knowledge, addressing API coverage gaps, and proposing new sub-tasks and benchmarks for more informative conversations.

Contribution

It introduces a framework for incorporating unstructured knowledge into dialogue systems and provides an augmented dataset with new sub-tasks and baseline models.

Findings

External knowledge improves response informativeness.

New dataset with out-of-API-coverage turns created.

Baseline models demonstrate the need for further research.

Abstract

Most prior work on task-oriented dialogue systems are restricted to a limited coverage of domain APIs, while users oftentimes have domain related requests that are not covered by the APIs. In this paper, we propose to expand coverage of task-oriented dialogue systems by incorporating external unstructured knowledge sources. We define three sub-tasks: knowledge-seeking turn detection, knowledge selection, and knowledge-grounded response generation, which can be modeled individually or jointly. We introduce an augmented version of MultiWOZ 2.1, which includes new out-of-API-coverage turns and responses grounded on external knowledge sources. We present baselines for each sub-task using both conventional and neural approaches. Our experimental results demonstrate the need for further research in this direction to enable more informative conversational systems.