Recommendation Method of Cross-language Computer Courses

Abstract

AbstractIn traditional course recommendation methods, courses are usually recommended through a single language content. However, learners can not be effectively recommended using current methods when they want to access cross-language course content. In computer science, foreign computer technology has first mover advantages, so cross-language recommendation of foreign courses is important for learners. But there is still comparatively little work on related recommendation methods, traditional recommendation methods are insufficient to address the above needs. To meet the above needs, we propose a cross-language computer course recommendation method in this paper. Our method can be effectively used in the cross-language course recommendation scenario to recommend the most relevant top-N English courses through the content description information of Chinese courses. Our experiment was conducted in computer course data from 15 universities and a number of Massive Open Online Courses (MOOCs). As a result, our experiment achieved the accuracy of 89%, the precision of 90%, the recall of 85% and the F1-score of 86%. The evaluation indexes reflect the effectiveness and feasibility of our method. Our method can also be applied to course recommendation in other subject fields. KeywordsCourse recommendation systemCross-languageContent-based recommendation