Recent research on eco-driving has paid more attention on human driving behavior and vehicle performance, but neglects the impacts imposed by dynamic, discrete and distributive traffic, which may result in partial and limited eco-driving suggestions. This paper presents a multi-factor integration based eco-driving optimization method for vehicles with same driving characteristics. It constructs driving characteristics by integrating human-vehicle-road multiple influencing factors with Internet of vehicles, and sorts out drivers with same driving characteristics using dimension-reducing clustering algorithm, and then classifies energy-saving grading based on fuzzy control algorithm. Finally, eco-driving suggestions are provided for users who need to improve fuel saving by combining linear variable weight theory with expert database of economical driving speed. Contrast test on fuel consumption of two vehicles with same driving characteristics was conducted using vehicle simulation software Carmaker. With the suggestions provided by the method, average fuel consumption is 24.8% less than the vehicle with no suggestions. Test results show that eco-driving suggestions provided by the method are comprehensive, impartial and objective, which is more beneficial to save fuel and reduce bad driving habits.