PARAN: Persona-Augmented Review ANswering system on Food Delivery Review Dataset

TL;DR

This paper introduces PARAN, a two-stage prompting framework that infers user personas from reviews to generate personalized responses in food delivery services, improving relevance and engagement without fine-tuning models.

Contribution

The work presents a novel persona-augmented prompting method that infers explicit and implicit user personas from review texts for personalized response generation.

Findings

Enhanced response relevance and personalization.

Improved diversity and semantic consistency.

Effective without model fine-tuning.

Abstract

Personalized review response generation presents a significant challenge in domains where user information is limited, such as food delivery platforms. While large language models (LLMs) offer powerful text generation capabilities, they often produce generic responses when lacking contextual user data, reducing engagement and effectiveness. In this work, we propose a two-stage prompting framework that infers both explicit (e.g., user-stated preferences) and implicit (e.g., demographic or stylistic cues) personas directly from short review texts. These inferred persona attributes are then incorporated into the response generation prompt to produce user-tailored replies. To encourage diverse yet faithful generations, we adjust decoding temperature during inference. We evaluate our method using a real-world dataset collected from a Korean food delivery app, and assess its impact on precision, diversity, and semantic consistency. Our findings highlight the effectiveness of persona-augmented prompting in enhancing the relevance and personalization of automated responses without requiring model fine-tuning.