Realization of an Adaptive Test Paper Generation Function based on DPC Algorithm

Before COVID-19, although the online assessment platform has developed, it is relatively slow, and people prefer to organize on-site examinations. After the outbreak of the epidemic, people realize the urgency and necessity of information construction in all walks of life. More and more researchers begin to pay attention to and explore the use of advanced machine learning methods to improve the practicability of online examinations platform. The test paper generation function is the core link, and a good test paper generation function can ensure the quality of a test paper. In this paper, an advanced unsupervised algorithm DPC(Clustering by Fast Search and Find of Density Peaks) is used to conduct deep mining and test paper generation adaptive based on the question bank and historical assessment such as assessment frequency, accuracy or score rate, and a more reasonable test paper generation function is realized. By comparing and testing the experimental results, it can be proved that the idea is correct and feasible.