Deepfake Characterization, Propagation, and Detection in Social Media - A Synthesis Review

Deepfake technology, encompassing the creation of hyper-realistic manipulated media, has emerged as a potent force in shaping information dissemination on social media platforms. This synthesis review navigates through existing literature to comprehensively characterize and analyze the propagation dynamics and evaluate the efficacy of detection mechanisms employed in deepfakes within social media environments. Motivated by the escalating concerns surrounding the societal impact of deepfakes, this review critically assesses the current state of research, identifying key trends and challenges. It synthesizes findings on the diverse techniques employed in deepfakes creation, the mechanisms driving their dissemination across social networks, and the evolving landscape of detection strategies. Our synthesis reveals existing research gaps, emphasizing the need for more comprehensive investigations into the characteristics, behavioral, and propagation patterns of different deepfake types. Additionally, we highlight the necessity for standardized datasets and benchmarks to facilitate the development and evaluation of detection methods. Our review aims to guide future research endeavors by providing an overview of the current understanding of deepfake technologies while pinpointing areas that require further exploration and refinement.