# Emotion-Cause Pair Extraction in Conversations via Semantic Decoupling and Graph Alignment

Source: [Emotion-Cause Pair Extraction in Conversations via Semantic Decoupling and Graph Alignment](<https://www.sota2.com/research/paper/69e85bf87d0172140ec812a2>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-23T05:00:12.895Z

Authors: Tianxiang Ma, Weijie Feng, Xinyu Wang, Zhiyong Cheng

Published: 2026-04-21T15:05:48.000Z

## Abstract

Emotion-Cause Pair Extraction in Conversations (ECPEC) aims to identify the set of causal relations between emotion utterances and their triggering causes within a dialogue. Most existing approaches formulate ECPEC as an independent pairwise classification task, overlooking the distinct semantics of emotion diffusion and cause explanation, and failing to capture globally consistent many-to-many conversational causality. To address these limitations, we revisit ECPEC from a semantic perspective and seek to disentangle emotion-oriented semantics from cause-oriented semantics, mapping them into two complementary representation spaces to better capture their distinct conversational roles. Building on this semantic decoupling, we naturally formulate ECPEC as a global alignment problem between the emotion-side and cause-side representations, and employ optimal transport to enable many-to-many and globally consistent emotion-cause matching. Based on this perspective, we propose a unified framework SCALE that instantiates the above semantic decoupling and alignment principle within a shared conversational structure. Extensive experiments on several benchmark datasets demonstrate that SCALE consistently achieves state-of-the-art performance. Our codes are released at https://github.com/CoCoSphere/SCALE.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Emotion-Cause Pair Extraction in Conversations on ECF](<https://www.sota2.com/research/sota/emotion-cause-pair-extraction-in-conversations-on-ecf.md>): Emotion-Cause Pair Extraction in Conversations; ECF; F1 Score: 57.7.

- [Emotion-Cause Pair Extraction in Conversations on RECCON-DD](<https://www.sota2.com/research/sota/emotion-cause-pair-extraction-in-conversations-on-reccon-dd.md>): Emotion-Cause Pair Extraction in Conversations; RECCON-DD; F1 Score: 58.83.

- [Emotion-Cause Pair Extraction in Conversations on RECCON-IE](<https://www.sota2.com/research/sota/emotion-cause-pair-extraction-in-conversations-on-reccon-ie.md>): Emotion-Cause Pair Extraction in Conversations; RECCON-IE; Precision: 42.54.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2604.19547>)

[PDF](<https://arxiv.org/pdf/2604.19547.pdf>)
