Multi-Sentence Knowledge Selection in Open-Domain Dialogue

TL;DR

This paper evaluates current open-domain dialogue knowledge selection methods, identifies flaws, and introduces WOW++, an augmented dataset with multiple relevant knowledge sentences, to improve knowledge ranking algorithms.

Contribution

It proposes a new framework for collecting relevant knowledge and creates WOW++, a dataset with multiple knowledge sentences per context, enhancing evaluation and training.

Findings

Neural rerankers using WOW++ outperform those trained on standard datasets.

Existing methodologies have flaws in data and evaluation for knowledge selection.

WOW++ averages 8 relevant knowledge sentences per dialogue, capturing ambiguity.

Abstract

Incorporating external knowledge sources effectively in conversations is a longstanding problem in open-domain dialogue research. The existing literature on open-domain knowledge selection is limited and makes certain brittle assumptions on knowledge sources to simplify the overall task (Dinan et al., 2019), such as the existence of a single relevant knowledge sentence per context. In this work, we evaluate the existing state of open-domain conversation knowledge selection, showing where the existing methodologies regarding data and evaluation are flawed. We then improve on them by proposing a new framework for collecting relevant knowledge, and create an augmented dataset based on the Wizard of Wikipedia (WOW) corpus, which we call WOW++. WOW++ averages 8 relevant knowledge sentences per dialogue context, embracing the inherent ambiguity of open-domain dialogue knowledge selection. We then benchmark various knowledge ranking algorithms on this augmented dataset with both intrinsic evaluation and extrinsic measures of response quality, showing that neural rerankers that use WOW++ can outperform rankers trained on standard datasets.