How to start a conversation proactively?
How to start a conversation proactively?
This paper addresses the challenge of proactive conversation initiation for social robots by proposing a schema of contextual variables and introducing a curated dataset (CCO) of contextually-aligned openers. The authors fine-tune small language models (SLMs) using this dataset and show through comparative evaluation that explicit contextual structuring significantly improves the quality and appropriateness of conversation initiations compared to zero-shot LLMs and off-the-shelf SLMs.
The authors introduce a schema of contextual variables and a curated dataset (CCO) to fine-tune small language models for proactive conversation initiation in embodied social robots, demonstrating improved contextual appropriateness over zero-shot LLMs and off-the-shelf SLMs.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
Source：arXiv Robotics — research abstracts · arxiv.org