Cognitive Cyber Crimes in the Era of Artificial Intelligence

11Algorithmic Bias and Psychographic Profiling for Manipulation

Hitesh Rawat¹* and Anand Rajavat²

¹Computer Science Department, Shri Vaishnav Vidyapeeth Vishwavidyalaya, Indore, India

²Department of Computer Science Engineering, Shri Vaishnav Institute of Information Technology, Shri Vaishnav Vidyapeeth Vishwavidyalaya, Indore, India

Abstract

This study investigates the intersection of algorithmic bias and psychographic profiling as tools for manipulation in digital environments. Utilizing the Cambridge Analytica Dataset provided by the Open Psychometrics Project—comprising psychometric profiles and user interaction data of 50,000 participants—our methodology combines fairness-aware gradient boosting machine (FairGBM) with behavioral clustering via hierarchical agglomerative clustering. Key advanced parameters include Bias Amplification Score, Manipulation Potential Index (MPI), and psychographic vulnerability coefficient. The model achieved a bias detection accuracy of 94.7%, an MPI increase of 37% in targeted manipulation scenarios, and reduced unfair treatment disparity by 23% compared to baseline algorithms. This research highlights critical ethical considerations and proposes a framework to mitigate manipulative bias in psychographic targeting systems.

Keywords: Algorithmic bias, psychographic profiling, manipulation, fairnessaware machine learning, behavioral clustering

11.1 Introduction

In the era of data-driven decision-making, algorithmic bias and psychographic profiling ...