Document Type : Original Article
Authors
1
Ph.D. in IT Engineering , Department of Computer Engineering and Information Technology(E-Commerce), University of Qom, Qom, Iran.
2
Department of Computer Engineering and Information Technology, University of Qom
3
Department of Information Technology Engineering (E-Commerce), Faculty of Engineering, University of Qom, Qom, Iran.
4
MSc in Computer Engineering – Artificial Intelligence, Malek Ashtar University of Technology, Tehran, Iran
Abstract
This study presents a semi-automated framework for constructing a structured association map between cognitive neuroscience and marketing using knowledge graphs, large language models (LLMs), and expert validation. An existing Cognitive Neuroscience Knowledge Graph (CNKG) containing 312 neuroscience concepts served as the source domain, while a curated set of 100 marketing concepts formed the target domain.
The mapping process consisted of two phases: (1) automated association generation, where four LLMs (GPT-4, Gemini, Claude, and Grok) inferred candidate relationships between domains, and (2) expert validation, where marketing and economics specialists refined and filtered the generated associations.
The automated phase produced 1,284 candidate relations. After expert validation, 742 cross-domain relationships were retained, forming a bipartite knowledge graph with 412 nodes and moderate connectivity (average node degree = 3.57). Quantitative evaluation showed that 53.3% of relations were proposed by at least two LLMs, indicating moderate inter-model agreement. Expert assessment resulted in 39.9% accepted relations, 17.9% modified relations, and 42.2% rejected relations, highlighting the necessity of human supervision for scientific rigor.
Graph analysis revealed that core neuroscience concepts—such as reward processing, emotion, attention, and decision-making—acted as hub nodes strongly connected to marketing constructs including brand loyalty, purchase intention, advertising effectiveness, and customer engagement.
The proposed association map provides a novel theoretical framework for neuromarketing by systematically linking neural mechanisms to marketing phenomena. This hybrid AI–human approach demonstrates an effective paradigm for building interdisciplinary knowledge bridges and offers practical support for hypothesis generation, education, and evidence-based marketing strategy design.
Keywords
Subjects
