A Comparative Analysis of Deep Learning Algorithms in Enhancing Knowledge Management Systems

Abstract

This study aims to comparatively examine the role and effectiveness of deep learning algorithms in improving the processes, performance, and efficiency of knowledge management systems in organizations. This qualitative review employed a systematic content analysis approach. Data were collected from peer-reviewed articles published between 2018 and 2023 in databases such as Scopus, Web of Science, and IEEE Xplore. A total of 12 articles were selected based on theoretical saturation and analyzed using NVivo version 14. The analysis process included open, axial, and selective coding to extract key patterns and concepts related to deep learning in knowledge management. The results revealed that deep learning algorithms, particularly Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), and Generative Adversarial Networks (GAN), play a critical role in the creation, organization, and sharing of knowledge. Transfer learning and hybrid deep models enhanced both accuracy and computational efficiency in knowledge management systems. The findings also indicated that deep learning effectively supports all stages of the knowledge management cycle—from acquisition to evaluation—though challenges such as data quality, computational cost, and model interpretability persist. Deep learning provides an innovative framework for transforming organizational knowledge management. By identifying hidden patterns and analyzing complex data, it improves knowledge-based decision-making and fosters self-learning systems. However, to fully exploit its potential, developing interpretable, ethical models and strengthening data infrastructure are essential.