AI-Driven Personalization and Privacy Issues: A Study of The Literature Using Citation Network Analysis

Abstract

 AI-based personalisation has become one of the key components of current technological solutions. Personalisation technologies allow for adjusting the content, services and recommendations according to the preferences of individual users. Although the collection and inference of personal information is extremely beneficial, it creates severe privacy concerns. The present study offers a citation network analysis (CNA) of 362 primary articles on AI-based personalisation and privacy, which are collected from Scopus and Web of Science sources and result in a citation network of 28,967 nodes and 31,114 directed links between 2013 and 2026. For the purposes of main path analysis, we employed Pajek software, while community detection was conducted using Gephi. Three main topics were identified, namely (1) AI-based marketing personalisation and consumer behaviour, (2) human-computer interaction, explainability and user trust, and (3) ethics, regulation and digital well-being. Four bridge nodes exist on the central route connecting the clusters. The node with the highest betweenness centrality score is 250. Kumar et al., (2024), on the other hand, is the most cited document in the corpus with 10 local citations. Based on the findings, the area can be considered only partly siloed since there seems to be a lack of integration between technical privacy solutions and regulatory discussion.