Don't Forget Your ABC's: Evaluating the State-of-the-Art in Chat-Oriented Dialogue Systems
This work addresses the lack of standardized evaluation for open-domain dialogue systems, which is a critical issue for researchers and developers aiming to reliably assess and improve these systems, though it is incremental in refining evaluation techniques.
The paper tackles the problem of high-variance and inconsistent human evaluation in chat-oriented dialogue systems by introducing a novel behavior-based method, which is shown to be more suitable than existing approaches like Likert-style or comparative methods for dimensional evaluation.
Despite tremendous advancements in dialogue systems, stable evaluation still requires human judgments producing notoriously high-variance metrics due to their inherent subjectivity. Moreover, methods and labels in dialogue evaluation are not fully standardized, especially for open-domain chats, with a lack of work to compare and assess the validity of those approaches. The use of inconsistent evaluation can misinform the performance of a dialogue system, which becomes a major hurdle to enhance it. Thus, a dimensional evaluation of chat-oriented open-domain dialogue systems that reliably measures several aspects of dialogue capabilities is desired. This paper presents a novel human evaluation method to estimate the rates of many dialogue system behaviors. Our method is used to evaluate four state-of-the-art open-domain dialogue systems and compared with existing approaches. The analysis demonstrates that our behavior method is more suitable than alternative Likert-style or comparative approaches for dimensional evaluation of these systems.