K-PERM: Personalized Response Generation Using Dynamic Knowledge Retrieval and Persona-Adaptive Queries

TL;DR

K-PERM is a novel personalized conversational agent that dynamically integrates user personas and external knowledge sources to generate more relevant and comprehensive responses, significantly improving personalization in dialogue systems.

Contribution

This paper introduces K-PERM, a dynamic approach combining persona adaptation and knowledge retrieval, achieving state-of-the-art results on personalized conversation benchmarks.

Findings

K-PERM outperforms previous models on the FoCus dataset.

Responses from K-PERM enhance GPT 3.5 performance by 10.5%.

The method improves relevance and personalization in conversational agents.

Abstract

Personalizing conversational agents can enhance the quality of conversations and increase user engagement. However, they often lack external knowledge to appropriately tend to a user's persona. This is particularly crucial for practical applications like mental health support, nutrition planning, culturally sensitive conversations, or reducing toxic behavior in conversational agents. To enhance the relevance and comprehensiveness of personalized responses, we propose using a two-step approach that involves (1) selectively integrating user personas and (2) contextualizing the response with supplementing information from a background knowledge source. We develop K-PERM (Knowledge-guided PErsonalization with Reward Modulation), a dynamic conversational agent that combines these elements. K-PERM achieves state-of-the-art performance on the popular FoCus dataset, containing real-world personalized conversations concerning global landmarks. We show that using responses from K-PERM can improve performance in state-of-the-art LLMs (GPT 3.5) by 10.5%, highlighting the impact of K-PERM for personalizing chatbots.