Application Research of Hierarchical Adaptive Online Course Teaching System based on Optimization Learning Algorithm

Accompanied by the continuous development of computer network technology, each university is eager to establish a set of scientific, reasonable, convenient, efficient and easy-to-use online course evaluation systems. This paper proposes an adaptive online assessment system for course teaching based on the adaptive learning algorithm with the improved hierarchical method. First, this paper combines the Hidden Markov Model and proposes a selection quantity reflecting the probability of items being selected according to the relationship between the exposure rate and the accuracy of the quiz, and puts forward a selection strategy based on the difficulty and differentiation of multilevel stratification, and controlling the exposure rate with the number of item subjects, i.e., the “b-a multilevel hierarchical selection method for controlling the exposure rate”. Then the adaptive online assessment system and its database are designed according to the optimized adaptive learning algorithm. The test results found that the b-a multilevel hierarchical question selection method proposed in this paper to control the exposure rate is effective in controlling the exposure rate relative to the maximum information selection, and the number of times the items are exposed is averaged through the computation of the exposure rate, but the exposure rate is not as good as the hierarchical method in the highest differentiation level of the hierarchical division as in the case of the informativeness hierarchical method, and the distribution of the exposure rate of the asymptotic maximum information method is the closest to that of the hierarchical method in terms of the exposure rate. Exposure is only second to the a-stratification method. The maximum response time of the general assessment system is 40[Formula: see text]s when virtualizing the transaction transmission of 10 users, while the maximum response time of the test system in this paper is 50[Formula: see text]s when virtualizing the transaction transmission of 50 users. This paper is based on an optimized adaptive learning algorithm of the course teaching test system, which improves the performance of the system with concurrent access, reduces the user waiting time and provides a well-performing online assessment system platform.