브랜드 언어로 본 글로벌 패션 스타트업: 텍스트 마이닝 기반 전략 키워드 비교 분석

With the rise of digital-native generations, launching an independent brand has emerged as a viable career path among fashion majors. In this context, the linguistic construction of brand identity has become an essential strategy, especially for early-stage fashion startups seeking to communicate their values to consumers, investors, and partners. This study aims to analyze how global fashion startups strategically express their brand identity through language, focusing on the official website texts, including About, Brand Story, and Mission Statement. The research examined 34 global fashion startups founded between 2013 and 2022. Text mining techniques, including frequency analysis, TF-IDF, 2-gram analysis, topic modeling (LDA), and hierarchical clustering based on cosine distance were applied using Orange3 and Python. Results revealed that while terms like material, design, and sustainable were common across brands, each category highlighted distinct keywords aligned with their identity and values. Topic modeling identified five dominant themes, including sustainability, artistic expression, and circularity. Cluster analysis classified brands into four language strategy groups, such as emotion-driven design brands and impact-oriented future-focused brands. These findings demonstrate that even in early-stage communication, fashion startups strategically design brand narratives through language to convey identity, build emotional connections, and distinguish themselves in a competitive market. This study contributes to understanding brand language as a core asset in fashion branding and offers a data-driven framework for analyzing identity-building in startup communication.