"Stupid robot, I want to speak to a human!" User Frustration Detection in Task-Oriented Dialog Systems

TL;DR

This paper evaluates various methods for detecting user frustration in real-world task-oriented dialog systems, highlighting the superiority of LLM-based approaches over open-source solutions with a 16% F1 score improvement.

Contribution

It compares multiple frustration detection methods in a deployed system and demonstrates the effectiveness of in-context learning LLMs over traditional open-source techniques.

Findings

LLM-based detection outperforms open-source methods by 16% F1 score

Open-source sentiment analysis methods have limitations in real-world frustration detection

The study provides industry insights into user frustration detection challenges

Abstract

Detecting user frustration in modern-day task-oriented dialog (TOD) systems is imperative for maintaining overall user satisfaction, engagement, and retention. However, most recent research is focused on sentiment and emotion detection in academic settings, thus failing to fully encapsulate implications of real-world user data. To mitigate this gap, in this work, we focus on user frustration in a deployed TOD system, assessing the feasibility of out-of-the-box solutions for user frustration detection. Specifically, we compare the performance of our deployed keyword-based approach, open-source approaches to sentiment analysis, dialog breakdown detection methods, and emerging in-context learning LLM-based detection. Our analysis highlights the limitations of open-source methods for real-world frustration detection, while demonstrating the superior performance of the LLM-based approach, achieving a 16\% relative improvement in F1 score on an internal benchmark. Finally, we analyze advantages and limitations of our methods and provide an insight into user frustration detection task for industry practitioners.