텍스트마이닝 분석기법을 활용한 AI 기반 금융서비스에 관한 특성요인 분석 연구

A Study on the Characteristic Factors of AI-Based Financial Services Using Text Mining Techniques

This study identifies the core quality characteristics of AI‑based financial services and analyzes them through the lens of the Information Systems Success Model (ISSM), which comprises system quality, information quality, and service quality. To this end, 150 academic papers and 318 news articles were collected and preprocessed, and a combination of LDA, BERTopic, and CONCOR was used to extract major quality attributes, which were then classified in relation to the features of intelligent financial platforms. The results indicate that AI services emphasize personalization grounded in usefulness, creativity, and interactivity, highlighting cognitive and affective user experience. Financial services are characterized by trust, accessibility, and empathy, particularly in non‑face‑to‑face channels and in the inclusion of digitally marginalized users. IT systems exhibit practicality, integration, and ease of use, underscoring interface design and technical reliability. Digital financial services, as a layered integration of these three domains, require strategies that balance technological trustworthiness with user‑centered trust. Academically, this study presents an integrated classification that systematically maps the three ISSM quality dimensions to the outputs of LDA, BERTopic, and CONCOR, thereby clarifying the conceptual foundations of quality research on AI‑based financial services. Practically, it prioritizes key quality factors trust, explainability, accuracy, accessibility, UX, and responsiveness and proposes design directions focused on non‑face‑to‑face channels, inclusion of vulnerable user groups, and enhanced transaction stability and security, offering actionable guidance for service strategy and quality improvement.