Fairness-Aware Recruitment Using Convolutional Neural Networks Combined with Support Vector Machines

Ensuring fairness in recruitment has become a critical challenge in modern organizations, particularly with the increasing adoption of Artificial Intelligence (AI) in decision-making. This paper proposes a hybrid framework that integrates Convolutional Neural Networks (CNNs) with Support Vector Machines (SVMs) to address bias in recruitment processes. The CNN component is employed for feature extraction from candidate profiles, including textual and image-based data, while the SVM is utilized for classification and fairness adjustment. To mitigate algorithmic bias, fairness-aware constraints are incorporated into the training phase, ensuring equitable outcomes across demographic groups. Experimental results on benchmark recruitment datasets demonstrate that the proposed CNN-SVM model significantly improves fairness metrics such as Demographic Party (DP) and Equal Opportunity (EO) without compromising classification accuracy. This approach highlights the potential of combining deep learning and traditional machine learning models to promote transparent, unbiased, and effective recruitment practices in real-world applications.

Keywords:

Fairness-aware recruitment , CNN , SVM , Bias mitigation , Demographic parity , Equal opportunity