Nowadays

the rapid adoption of electric vehicles (EVs) across modern countries has caused significant challenges in EV management in modern transportation systems. This creates a significant demand for efficient and intelligent recommendation systems to decrease EVs' travel time and cost and minimise charging wait time. Recently

many researchers have addressed this issue by designing AI-based EV charging station (EVCS) recommendation systems. In contrast to traditional routing problems

EVCS recommendations must consider dynamic demand that shifts across time and location and diverse infrastructure constraints. This survey article first introduces a categorisation of EVs and charging stations. Then

we provide a comprehensive review of different EVCS recommendation systems

including algorithmic and optimization frameworks

and recent deep learning (DL)

multi-agent reinforcement learning (MARL)

federated learning (FL)

and blockchain approaches. Furthermore

this survey provides a unified problem formulation for EVCS recommendation systems and examines the performance of different recommendation categories in terms of reliability

cost-saving

and availability scores. Moreover

this paper reviews the characteristics of publicly available datasets in this realm and provides future research directions for designing EVCS intelligent recommendation systems. Experimental results on the UrbanEV dataset show the superior performance of MARL methods compared with other baselines in terms of the considered key performance metrics.