SYNTHEMPATHY introduces a novel framework for creating large-scale empathetic dialogue datasets using LLMs, eliminating the need for expensive human annotation.

• Automated generation pipeline that produces high-quality empathetic conversations
• Cost-effective scaling compared to traditional crowdsourcing approaches
• Comparable quality to human-created datasets, as validated through evaluation
• Applicable across domains including customer support, healthcare, and general conversations

This research enables support teams to built more empathetic AI systems without the traditional costs and time constraints of data collection, potentially improving customer satisfaction and agent effectiveness in emotional scenarios.

SYNTHEMPATHY: A Scalable Empathy Corpus Generated Using LLMs Without Any Crowdsourcing