Steam flooding is an important thermal recovery method for heavy oil reservoirs, and convective heating technology is used to fracture oil shale reservoirs with good results. This paper reviews the main prediction methods, optimization approaches for steam flooding performance, and its application in fractured oil shale reservoirs. The prediction methods include experimental, numerical simulation, and statistical models. These provide insights into steam override, heat transfer, and production dynamics. To optimize steam flooding, parameters like quality, temperature, injection rate, and allocation need to be coordinated based on reservoir conditions and monitoring data. Real-time injection control, economic analysis, and sweep efficiency improvements should also be considered in optimization workflows. Although progress has been made, more field studies are needed to establish systematic optimization practices utilizing advanced technologies. This review summarizes the key developments in steam flooding modeling and optimization, providing a reference for further research and field applications.