Analysis And Design Of Evaluation System Of Online Peer Reviewers Based On Information Integration Method

The rapidly increased online courses provide convenience for learners to study online and among these online courses the massive open online course (MOOC) is prominent. In order to deal with the problem of reviewing large numbers of homework in a short period of time, the MOOC designers introduce the peer assessment process into online courses. Peer assessment in MOOC plays an important role in reducing teachers’burden, enriching learner’s knowledge, broadening learner’s views and raising the enthusiasm of learners to participate in the online curriculum. However, many MOOC learners point out that the existing online peer assessment cannot generate satisfactory result in assigning appropriate peer reviewers for submissions. The problem is expressed as the rude review comments and the too high or too low homework scores. Therefore, this paper proposes a peer reviewer recommendation approach to match reviewers and submissions under certain constrains. The approach integrates submissions’information and reviewers’ information from three dimensions:reviewer intention, reviewer assessment ability and the relevance between reviewers and submissions. To test the effectiveness of this approach, an experiment on 46 college students was conducted to test the effect of the peer reviewer recommendation approach. The experiment shows that the proposed recommendation approach can generate more satisfactory results than baseline approach in the following aspects: submissions’final score, submissions’ comments, peer reviewers’attitude and comparing the final score with the score which is given by tutors. Therefore, our peer reviewer recommendation plays an obvious role in improving learners’ satisfaction of peer assessment process.