Machine learning-based research of AI marketing: topic analysis and model construction

Abstract

Artificial intelligence (AI) has emerged as a transformative force in the marketing landscape, significantly altering consumer engagement, decision-making processes, and strategic management practices. However, despite the substantial expansion of research on AI in marketing, a comprehensive and systematic understanding of its underlying knowledge structure and the evolution of its thematic domains remains insufficient. Therefore, we synthesize qualitative insights with bibliometric science mapping by applying CiteSpace and latent Dirichlet allocation (LDA) topic modeling to 564 Web of Science Core Collection articles (2019–May 2025). The analysis yields three core findings. First, China and the USA anchor the field’s knowledge production, with the UK, Australia, and India as secondary hubs. Second, LDA reveals five topics: (1) Digitalization Transformation and Marketing Strategy, which explores the impact of digital technologies on reshaping traditional marketing approaches; (2) Consumer Behavior and Experience, focusing on how AI influences consumer decision-making and engagement; (3) Business Value and Innovation Research, examining how AI contributes to business value creation and the development of innovative marketing practices; (4) Machine Learning Algorithms and Technological Applications, addressing the technical foundations and practical implementations of AI in marketing; and (5) Generative AI Content and Applications, highlighting the emerging role of generative AI in content creation and personalization. Third, we develop an integrative value chain framework that explicates how AI permeates and transforms distinct stages of the marketing process, offering a holistic perspective on AI’s role in modern marketing practices. Building on these results, we advance a dual co-evolutionary framework—Consumer–Technology–Market and Technology–Economy–Society—that consolidates dispersed streams, clarifies the field’s intellectual structure. For managers, we map the AI–marketing technology landscape, identify high-yield opportunities in customer experience and service automation, and delineate risks involving data privacy, bias, and ethical guardrails, thereby informing effective and responsible adoption of AI in marketing strategy.