Related Experiment Video
Updated: Jun 30, 2026

Integrating Computerized Linguistic and Social Network Analyses to Capture Addiction Recovery Capital in an Online Community
Published on: May 31, 2019
Integrating NLP and expert validation: a framework combining subjective and objective approaches for female-oriented
1Department of Design, Graduate School, Hanyang University, Seoul, 04763, Republic of Korea.
Abstract:
This study responds to the rapid expansion of the female automobile consumer market and the corresponding lack of detailed user insight by proposing a hybrid methodology that combines natural language processing (NLP) with expert workshops. The primary objective is to generate refined personas of female car users based on social media data, thereby offering actionable insights to guide automotive design and marketing strategies and to elucidate the psychological and behavioral mechanisms influencing women's perceptions and purchasing decisions. A total of 273,657 content entries from 12,866 female users were collected across four major platforms-Weibo, Xiaohongshu, Bilibili, and Autohome. A four-stage analytical framework was employed. First, structured data were quantitatively clustered using the Latent Dirichlet Allocation (LDA) topic model and the Qwen-3 large language model, resulting in the identification of five representative user groups with distinct behavioral characteristics. These clusters were subsequently refined through interdisciplinary expert workshops to construct semantically coherent and narratively rich personas. Each persona integrates demographic attributes, vehicle usage contexts, and emotional preferences. To assess the reliability of the generated personas, the study introduces a six-dimensional Persona Perception Scale (PPS), which confirmed the personas' high credibility. This research contributes to the field in three key ways: (1) by establishing a multi-platform data fusion framework to bridge the gap between raw data and narrative user modeling; (2) by developing an LLM-compatible algorithmic pipeline for fine-grained need extraction; and (3) by proposing a scalable and replicable model for data-driven persona generation. The methodology offers a validated, empirical tool for targeted marketing and a deeper understanding of the decision-making pathways of female automobile consumers.
More Related Videos
07:14Virtual Agent for Real-Time Motivational Interviewing by Integrating Adaptive Nonverbal Behavior and Language Models
Published on: December 23, 2025
09:53Humor or Rationality? The Neural Mechanisms of How Agent Type and Language Style Influence Satisfaction with Ride-Hailing Service Failure Recovery
Published on: March 13, 2026
Related Concept Videos
Social Proof
Stereotype Content Model
Internal Combustion Engine
Design Example: Automobile Ignition System
One can generate a large voltage using a car battery of 12 volts with the help of inductors. Inductors are known for opposing...
Impression Management Techniques IV: Altercasting
Automatic Processing and Automatic Social Behavior