The use of Electric Vehicles (EVs) is increasing rapidly across the world, making accurate battery state-of-charge (SoC) estimation essential for safe, reliable, and efficient operation. Since SoC cannot be directly measured, it must be inferred from observable values such as voltage, current, and temperature. Data-driven methods have recently emerged as a powerful method for this estimation task. This paper presents a focused review of data-driven SoC estimation techniques published over the past five years, categorizing existing approaches into temporal-based, tree-based, spatial-based, and edge-oriented models. Commonly used datasets at both cell and pack levels are also surveyed. Finally, a gap analysis is conducted to identify current limitations and highlight open research challenges and future directions.