Partiality for one or some subject(s) is an important issue in higher education and may directly affect the learning outcome of students. In this paper, we will calculate the difference between individual learning outcome and average learning outcome with collaborative filtering of personalized recommendation based on the binary relation between learning outcome of a student and average learning outcome of all students, and use the calculation result to find the students who are partial to one or some subject(s) and need help. Then personalized recommendation and feedback will be given to such students according to their specific situations in order to help them learn with a more effective method and have a positive impact on their learning outcomes in future. This method can assist colleges and universities in management and control of learning process of students. And we find this method can also be used to find the students partial on one or some subject(s) effectively and accurately.