Exploring the Frontier: Generative AI Applications in Online Consumer Behavior Analytics

Abstract

This paper presents a systematic review of the application of generative artificial intelligence (AI) in online consumer behavior analytics (OCBA). With the advent of e-commerce and social media, consumer behavior increasingly occurs online, generating vast amounts of data. This shift necessitates advanced analytical tools, and generative AI emerges as a pivotal technology. Generative AI, distinct from traditional AI, can autonomously generate new content based on learned data patterns, offering innovative approaches to OCBA. Based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) methodology and data synthesis method proposed by Webster and Watson (2002), this study analyzes 28 peer-reviewed papers, focusing on how generative AI is applied in OCBA and how it can enhance OCBA performance. The findings show that generative adversarial networks (GANs) are the most used, followed by variational autoencoders (VAEs) and autoregressive models. This review categorizes the application areas of generative AI in OCBA and examines how these technologies enhance OCBA’s effectiveness and efficiency. Furthermore, the paper discusses the challenges associated with generative AI, emphasizing the need to consider ethical issues, such as bias and data privacy. This comprehensive review contributes to a deeper understanding of generative AI’s role in OCBA, outlining its applications and functionalities from a technical perspective. It guides future research and practice, highlighting areas for further exploration and improvement in leveraging generative AI for consumer behavior analytics.

Keywords

• generative artificial intelligence
• generative adversarial network
• variational autoencoders
• generative pre-trained transformer
• online consumer behavior analytics
• autoregressive model