Navigating the Unknown: A Chat-Based Collaborative Interface for Personalized Exploratory Tasks

Yingzhe Peng,Xiaoting Qin,Zhiyang Zhang,Jue Zhang,Qingwei Lin,Xu Yang,Dongmei Zhang,Saravan Rajmohan,Qi Zhang

2024-10-31

Abstract:The rise of large language models (LLMs) has revolutionized user interactions with knowledge-based systems, enabling chatbots to synthesize vast amounts of information and assist with complex, exploratory tasks. However, LLM-based chatbots often struggle to provide personalized support, particularly when users start with vague queries or lack sufficient contextual information. This paper introduces the Collaborative Assistant for Personalized Exploration (CARE), a system designed to enhance personalization in exploratory tasks by combining a multi-agent LLM framework with a structured user interface. CARE's interface consists of a Chat Panel, Solution Panel, and Needs Panel, enabling iterative query refinement and dynamic solution generation. The multi-agent framework collaborates to identify both explicit and implicit user needs, delivering tailored, actionable solutions. In a within-subject user study with 22 participants, CARE was consistently preferred over a baseline LLM chatbot, with users praising its ability to reduce cognitive load, inspire creativity, and provide more tailored solutions. Our findings highlight CARE's potential to transform LLM-based systems from passive information retrievers to proactive partners in personalized problem-solving and exploration.

Human-Computer Interaction,Artificial Intelligence,Computation and Language

What problem does this paper attempt to address?

The paper attempts to address the issue of chatbots built on large language models (LLMs) struggling to provide personalized support when handling exploratory tasks. Specifically, when users start with vague queries or lack sufficient background information, existing LLM chatbots often fail to generate highly targeted and practical suggestions. For example, in travel planning, if a user only provides a vague query like "plan a 5-day trip to Hawaii" without specific budget, activity preferences, or accommodation requirements, the chatbot typically generates overly generic and impractical suggestions.

To tackle this challenge, the paper introduces a system called "Collaborative Assistant for Personalized Exploration" (CARE). CARE aims to enhance personalized support in exploratory tasks by combining a multi-agent LLM framework with a structured user interface. The system's user interface includes three main components: the Chat Panel, the Solution Panel, and the Needs Panel, which support iterative query refinement and dynamic solution generation. The design of these panels helps reduce the user's cognitive load, stimulate creative thinking, and provide more personalized solutions.

Overall, the goal of the CARE system is to transform LLMs from passive information retrieval tools into active partners in personalized problem-solving and exploration. Through a multi-agent collaborative framework, CARE can identify both explicit and implicit user needs, generating customized and actionable solutions, thereby enhancing the user experience.