PaRT: Enhancing Proactive Social Chatbots with Personalized Real-Time Retrieval

2025-04-29

AI Summary

Traditional social chatbots rely on user-initiated interactions, resulting in low engagement and limited dialogue duration. To address this, we propose an end-to-end proactive dialogue framework—the first to jointly model user profiles and dialogue context via intent refinement and real-time RedNote knowledge retrieval, enabling personalized topic generation and knowledge-augmented responses. Our approach integrates LLM-based prompt engineering, intent recognition, retrieval-augmented generation (RAG), and RedNote knowledge base integration, moving beyond passive response paradigms. Deployed in production for over 30 days, the system achieves a 21.77% average increase in dialogue length, significantly improving user retention and interaction depth. Key contributions include: (1) the first real-time proactive dialogue architecture designed specifically for social scenarios; and (2) a lightweight, intent-driven RAG paradigm enhanced by RedNote retrieval.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Natural Language Processing: Conversational AI/Dialog Systems

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Social chatbots have become essential intelligent companions in daily scenarios ranging from emotional support to personal interaction. However, conventional chatbots with passive response mechanisms usually rely on users to initiate or sustain dialogues by bringing up new topics, resulting in diminished engagement and shortened dialogue duration. In this paper, we present PaRT, a novel framework enabling context-aware proactive dialogues for social chatbots through personalized real-time retrieval and generation. Specifically, PaRT first integrates user profiles and dialogue context into a large language model (LLM), which is initially prompted to refine user queries and recognize their underlying intents for the upcoming conversation. Guided by refined intents, the LLM generates personalized dialogue topics, which then serve as targeted queries to retrieve relevant passages from RedNote. Finally, we prompt LLMs with summarized passages to generate knowledge-grounded and engagement-optimized responses. Our approach has been running stably in a real-world production environment for more than 30 days, achieving a 21.77% improvement in the average duration of dialogues.

Problem

• Enhancing chatbot engagement through proactive dialogue initiation
• Personalizing responses using real-time retrieval and user profiles
• Improving dialogue duration with context-aware topic generation

Innovation

• LLM integrates user profiles and dialogue context
• Personalized topics guide real-time retrieval from RedNote
• Knowledge-grounded responses enhance engagement and duration