TD-EVAL: Revisiting Task-Oriented Dialogue Evaluation by Combining Turn-Level Precision with Dialogue-Level Comparisons

• URL: http://arxiv.org/abs/2504.19982v1
• Date: Mon, 28 Apr 2025 16:57:17 GMT
• Title: TD-EVAL: Revisiting Task-Oriented Dialogue Evaluation by Combining Turn-Level Precision with Dialogue-Level Comparisons
• Authors: Emre Can Acikgoz, Carl Guo, Suvodip Dey, Akul Datta, Takyoung Kim, Gokhan Tur, Dilek Hakkani-Tür,
• Abstract summary: TD-EVAL (Turn and Dialogue-level Evaluation) is a two-step evaluation framework that unifies fine-grained turn-level analysis with holistic dialogue-level comparisons.<n>We show that TD-EVAL effectively identifies the conversational errors that conventional metrics miss.<n>It also exhibits better alignment with human judgments than traditional and Large Language Models-based metrics.
• Score: 11.961955016373379
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Task-oriented dialogue (TOD) systems are experiencing a revolution driven by Large Language Models (LLMs), yet the evaluation methodologies for these systems remain insufficient for their growing sophistication. While traditional automatic metrics effectively assessed earlier modular systems, they focus solely on the dialogue level and cannot detect critical intermediate errors that can arise during user-agent interactions. In this paper, we introduce TD-EVAL (Turn and Dialogue-level Evaluation), a two-step evaluation framework that unifies fine-grained turn-level analysis with holistic dialogue-level comparisons. At turn level, we evaluate each response along three TOD-specific dimensions: conversation cohesion, backend knowledge consistency, and policy compliance. Meanwhile, we design TOD Agent Arena that uses pairwise comparisons to provide a measure of dialogue-level quality. Through experiments on MultiWOZ 2.4 and {\tau}-Bench, we demonstrate that TD-EVAL effectively identifies the conversational errors that conventional metrics miss. Furthermore, TD-EVAL exhibits better alignment with human judgments than traditional and LLM-based metrics. These findings demonstrate that TD-EVAL introduces a new paradigm for TOD system evaluation, efficiently assessing both turn and system levels with a plug-and-play framework for future research.

Related papers

• Measuring the Robustness of Reference-Free Dialogue Evaluation Systems [12.332146893333952]
  We present a benchmark for evaluating the robustness of reference-free dialogue metrics against four categories of adversarial attacks.<n>We analyze metrics such as DialogRPT, UniEval, and PromptEval across grounded and ungrounded datasets.
  arXiv  Detail & Related papers  (2025-01-12T06:41:52Z)
• Don't Forget Your ABC's: Evaluating the State-of-the-Art in Chat-Oriented Dialogue Systems [12.914512702731528]
  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.
  arXiv  Detail & Related papers  (2022-12-18T22:07:55Z)
• FineD-Eval: Fine-grained Automatic Dialogue-Level Evaluation [58.46761798403072]
  We propose a dialogue-level metric that consists of three sub-metrics with each targeting a specific dimension. The sub-metrics are trained with novel self-supervised objectives and exhibit strong correlations with human judgment for their respective dimensions. Compared to the existing state-of-the-art metric, the combined metrics achieve around 16% relative improvement on average.
  arXiv  Detail & Related papers  (2022-10-25T08:26:03Z)
• DynaEval: Unifying Turn and Dialogue Level Evaluation [60.66883575106898]
  We propose DynaEval, a unified automatic evaluation framework. It is capable of performing turn-level evaluation, but also holistically considers the quality of the entire dialogue. Experiments show that DynaEval significantly outperforms the state-of-the-art dialogue coherence model.
  arXiv  Detail & Related papers  (2021-06-02T12:23:18Z)
• Towards Quantifiable Dialogue Coherence Evaluation [126.55560816209756]
  Quantifiable Dialogue Coherence Evaluation (QuantiDCE) is a novel framework aiming to train a quantifiable dialogue coherence metric. QuantiDCE includes two training stages, Multi-Level Ranking (MLR) pre-training and Knowledge Distillation (KD) fine-tuning. Experimental results show that the model trained by QuantiDCE presents stronger correlations with human judgements than the other state-of-the-art metrics.
  arXiv  Detail & Related papers  (2021-06-01T14:11:17Z)
• Assessing Dialogue Systems with Distribution Distances [48.61159795472962]
  We propose to measure the performance of a dialogue system by computing the distribution-wise distance between its generated conversations and real-world conversations. Experiments on several dialogue corpora show that our proposed metrics correlate better with human judgments than existing metrics.
  arXiv  Detail & Related papers  (2021-05-06T10:30:13Z)
• Modelling Hierarchical Structure between Dialogue Policy and Natural Language Generator with Option Framework for Task-oriented Dialogue System [49.39150449455407]
  HDNO is an option framework for designing latent dialogue acts to avoid designing specific dialogue act representations. We test HDNO on MultiWoz 2.0 and MultiWoz 2.1, the datasets on multi-domain dialogues, in comparison with word-level E2E model trained with RL, LaRL and HDSA.
  arXiv  Detail & Related papers  (2020-06-11T20:55:28Z)
• Rethinking Dialogue State Tracking with Reasoning [76.0991910623001]
  This paper proposes to track dialogue states gradually with reasoning over dialogue turns with the help of the back-end data. Empirical results demonstrate that our method significantly outperforms the state-of-the-art methods by 38.6% in terms of joint belief accuracy for MultiWOZ 2.1.
  arXiv  Detail & Related papers  (2020-05-27T02:05:33Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.