As a non-teaching factor, the influence of teachers can often trigger positive teaching effects, and exert impacts on students’ learning engagement. However, few of the existing studies on the online learning engagement of college students have viewed this topic from a quantitative perspective, and few have concerned about the influence mechanism of Teacher Influence (TI). To fill in this research blank, this paper studied the influence mechanism of TI on the learning engagement of students. At first, this paper analyzed the theories related to the said influence mechanism and proposed a TI model for the online teaching environment. Then, after investigating students’ understandings, participation, and questions with the online teaching activities created by teachers, the TI model was built in the paper. In view of the complexity of this TI model, this paper fully considered the meaning of the interactions of teacher and student nodes during the actual online teaching activities, proposed a two-stage TI maximization algorithm based on the learning input responses of the nodes, and realized comprehensive quantification of TI. The effectiveness of the proposed algorithm was verified by experimental results.