Prerequisite-aware course ordering towards getting relevant job opportunities

Abstract

Adapting learning experience according to the rapidly-changing job market is essential for students to achieve fruitful learning and successful career development. As building blocks of potential job opportunities, we focus on “technical terminologies” which are frequently required in the job market. Given a technical terminology, we aim at identifying an order of courses which contributes to the acquisition of knowledge about the terminology and also follows the prerequisite relationships among courses. To solve the course ordering problem, we develop a two-step approach, in which course-terminology relatedness is first estimated and then courses are ordered based on the prerequisite relationships and the estimated relatedness. Focusing on the second step, we propose a method based on Markov decision process (MDPOrd) and compare it with three other methods: (1) a method that orders courses based on aggregated relatedness (AggRelOrd), (2) a method that topologically sorts the courses based on personalized PageRank values (PageRankTS), and (3) a method that greedily picks courses based on the average relatedness (GVPickings). In addition to evaluating how the order prioritizes the related courses, we also evaluate from pedagogical perspectives, namely, how the order prioritizes specifically/generally fundamental courses, and how it places courses close to their prerequisites. Experimental results on two course sets show that MDPOrd outperforms the other methods in prioritizing related courses. In addition, MDPOrd is effective in ordering courses close to their prerequisites, but does not work well in highly ranking fundamental courses in the order.