The paper summarizes identification methods for research fronts, including the method based on bibliometrics and the automatic and semi-automatic methods based on computers. Their respective advantages and disadvantages are noted and it is suggested that such tools as semantic network should be utilized to identify research fronts in a deeper and more accurate manner.