OPERA: Harmonizing Task-Oriented Dialogs and Information Seeking Experience

TL;DR

This paper introduces OPERA, a unified model for task-oriented dialogs and information seeking, combining external knowledge sources to improve conversational AI's ability to complete tasks and answer questions.

Contribution

It proposes the new OB-TOD task, creates the OB-MultiWOZ dataset, and develops OPERA, a model integrating explicit and implicit knowledge for enhanced dialog capabilities.

Findings

OPERA outperforms baselines in task completion and QA accuracy.

Both explicit and implicit knowledge sources significantly improve performance.

The dataset enables training models that handle combined TOD and QA tasks.

Abstract

Existing studies in conversational AI mostly treat task-oriented dialog (TOD) and question answering (QA) as separate tasks. Towards the goal of constructing a conversational agent that can complete user tasks and support information seeking, it is important to build a system that handles both TOD and QA with access to various external knowledge. In this work, we propose a new task, Open-Book TOD (OB-TOD), which combines TOD with QA task and expand external knowledge sources to include both explicit knowledge sources (e.g., the Web) and implicit knowledge sources (e.g., pre-trained language models). We create a new dataset OB-MultiWOZ, where we enrich TOD sessions with QA-like information seeking experience grounded on external knowledge. We propose a unified model OPERA (Open-book End-to-end Task-oriented Dialog) which can appropriately access explicit and implicit external knowledge to tackle the defined task. Experimental results demonstrate OPERA's superior performance compared to closed-book baselines and illustrate the value of both knowledge types.