Towards Proactive Interactions for In-Vehicle Conversational Assistants Utilizing Large Language Models

TL;DR

This paper explores how large language models can enhance proactive interactions in in-vehicle conversational assistants, improving safety and user experience through a new framework and strategy validated by experiments.

Contribution

It introduces a novel framework with five proactivity levels and a 'Rewrite + ReAct + Reflect' strategy to improve LLM-based IVCA interactions, addressing user intent and context awareness.

Findings

LLMs outperform state-of-the-art models in success rate

Proactive level with strong assumptions and user confirmation is most effective

Framework and strategy are validated through feasibility and subjective experiments

Abstract

Research demonstrates that the proactivity of in-vehicle conversational assistants (IVCAs) can help to reduce distractions and enhance driving safety, better meeting users' cognitive needs. However, existing IVCAs struggle with user intent recognition and context awareness, which leads to suboptimal proactive interactions. Large language models (LLMs) have shown potential for generalizing to various tasks with prompts, but their application in IVCAs and exploration of proactive interaction remain under-explored. These raise questions about how LLMs improve proactive interactions for IVCAs and influence user perception. To investigate these questions systematically, we establish a framework with five proactivity levels across two dimensions-assumption and autonomy-for IVCAs. According to the framework, we propose a "Rewrite + ReAct + Reflect" strategy, aiming to empower LLMs to fulfill the specific demands of each proactivity level when interacting with users. Both feasibility and subjective experiments are conducted. The LLM outperforms the state-of-the-art model in success rate and achieves satisfactory results for each proactivity level. Subjective experiments with 40 participants validate the effectiveness of our framework and show the proactive level with strong assumptions and user confirmation is most appropriate.