# Co-persona: Leveraging LLMs and Expert Collaboration to Understand User Personas through Social Media Data Analysis

arXiv: https://arxiv.org/abs/2506.18269 · HTML: https://mlsanity.com/paper/2506.18269/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Jun 2025.

## Built on by (most cited first, 2 of 2)

- [POPI: Personalizing LLMs via Optimized Natural Language Preference Inference](https://mlsanity.com/paper/2510.17881/) (2025)
- [Synthia: Scalable Grounded Persona Generation from Social Media Data](https://mlsanity.com/paper/2507.14922/) (2025)

## Abstract (arXiv)

This study introduces Co-Persona, a methodological framework bridging large-scale social media analysis with authentic user understanding through systematic integration of Large Language Models and expert validation. Through a case study of B.Co, a Chinese manufacturer, we investigated Co-Persona application in bedside lamp development. Our methodology analyzed over 38 million posts from Xiao Hongshu, employing multi-stage data processing combining advanced NLP with expert validation. Analysis revealed five user personas derived from bedtime behaviors: Health Aficionados, Night Owls, Interior Decorators, Child-care Workers, and Workaholics-each showing unique pre-sleep activities and product preferences. Findings demonstrate Co-Persona enhances manufacturers' ability to process large datasets while maintaining user understanding. The methodology provides structured approaches for targeted marketing and product strategies. Research contributes to theoretical understanding of data-driven persona development and practical applications in consumer-driven innovation. Code and data available at https://github.com/INFPa/LLMwithPersona.