Online Judge System Topic Classification

Abstract

with the development of education, the program design teaching is receiving more and more attention as one of the core computer science courses in recently years, more and more schools are training students in conjunction with Online Judge Systems (OJs). A great number of OJ platforms are developed in domestic and foreign, and there are many suits of exercises and solutions reported in the websites. However, the OJs do not classify neither the knowledge points, nor the difficulty of these topics. Moreover, corresponding problem solving methods have not been organized and used by programming enthusiasts. In this paper, we designed a classification method to solve this problem by predicting the topic categories, conjecting the free topics and the solution resources, and classifying the unknown categories and the difficulty level of new topics. It will benefit students and teachers in related learning with OJs.