dorsal/arxiv
View SchemaEmplifAI: a Fine-grained Dataset for Japanese Empathetic Medical Dialogues in 28 Emotion Labels
| Authors | Wan Jou She, Lis Kanashiro Pereira, Fei Cheng, Sakiko Yahata, Panote Siriaraya, Eiji Aramaki |
|---|---|
| Categories | |
| ArXiv ID | 2601.10033vv1 |
| URL | https://arxiv.org/abs/2601.10033 |
| License | http://creativecommons.org/licenses/by/4.0/ |
Abstract
This paper introduces EmplifAI, a Japanese empathetic dialogue dataset designed to support patients coping with chronic medical conditions. They often experience a wide range of positive and negative emotions (e.g., hope and despair) that shift across different stages of disease management. EmplifAI addresses this complexity by providing situation-based dialogues grounded in 28 fine-grained emotion categories, adapted and validated from the GoEmotions taxonomy. The dataset includes 280 medically contextualized situations and 4125 two-turn dialogues, collected through crowdsourcing and expert review. To evaluate emotional alignment in empathetic dialogues, we assessed model predictions on situation--dialogue pairs using BERTScore across multiple large language models (LLMs), achieving F1 scores of 0.83. Fine-tuning a baseline Japanese LLM (LLM-jp-3.1-13b-instruct4) with EmplifAI resulted in notable improvements in fluency, general empathy, and emotion-specific empathy. Furthermore, we compared the scores assigned by LLM-as-a-Judge and human raters on dialogues generated by multiple LLMs to validate our evaluation pipeline and discuss the insights and potential risks derived from the correlation analysis.
{
"annotation_id": "0ed8a6bc-a1c2-480d-8fef-ef43ff9c9021",
"date_created": "2026-02-17T05:53:23.556000Z",
"date_modified": "2026-02-17T05:53:23.556000Z",
"file_hash": "de41a17fd6b4e67be974e99df2b45b0dc5f88d60ee4fc29d3ec6a7ed11fa88e0",
"private": false,
"record": {
"abstract": "This paper introduces EmplifAI, a Japanese empathetic dialogue dataset designed to support patients coping with chronic medical conditions. They often experience a wide range of positive and negative emotions (e.g., hope and despair) that shift across different stages of disease management. EmplifAI addresses this complexity by providing situation-based dialogues grounded in 28 fine-grained emotion categories, adapted and validated from the GoEmotions taxonomy. The dataset includes 280 medically contextualized situations and 4125 two-turn dialogues, collected through crowdsourcing and expert review. To evaluate emotional alignment in empathetic dialogues, we assessed model predictions on situation--dialogue pairs using BERTScore across multiple large language models (LLMs), achieving F1 scores of 0.83. Fine-tuning a baseline Japanese LLM (LLM-jp-3.1-13b-instruct4) with EmplifAI resulted in notable improvements in fluency, general empathy, and emotion-specific empathy. Furthermore, we compared the scores assigned by LLM-as-a-Judge and human raters on dialogues generated by multiple LLMs to validate our evaluation pipeline and discuss the insights and potential risks derived from the correlation analysis.",
"arxiv_id": "2601.10033",
"authors": [
"Wan Jou She",
"Lis Kanashiro Pereira",
"Fei Cheng",
"Sakiko Yahata",
"Panote Siriaraya",
"Eiji Aramaki"
],
"categories": [
"cs.CL"
],
"license": "http://creativecommons.org/licenses/by/4.0/",
"title": "EmplifAI: a Fine-grained Dataset for Japanese Empathetic Medical Dialogues in 28 Emotion Labels",
"url": "https://arxiv.org/abs/2601.10033",
"version": "v1"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "071c0564-9a17-47ad-92da-9ebf579edfa3",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}