| dc.description.abstract | With the current popularity of network technology, e-learning become more and more widely used. In recent years, Flipped Classroom, MOOC (Massive Open Online Course) and other learning mode are becoming the innovative study type.
Under this situation, there will be a large number of students to practice in a same course, it make teachers have the huge burden that they should judging a lot of students’ assignment. In order to solve the above-mentioned problem, this study presents a new peer assessment way which is consider students’ social relations first, student’s interaction information get from the online discussion board, and then, assign appropriate assessment for the student. Algorithm proposed in this study is designed to enhance the validity of peer assessment results. Meanwhile, in this study also explored the relationship between peer assessment accuracy and the students′ abilities.
Through the experimental results can be learned that the algorithm proposed in this study can effectively reduce the fairness issue in peer assessment. At the same time, it also verified the conclusion that student who have better ability can give the more accurate assessment result. From peer assessment results and teaching assistant assessment results correlation analysis results can get the significantly relevant conclusions, also verified the effectiveness of peer assessment in this study. Through questionnaire results, students for peer assessment activity hold the positive attitude. Description that students believe the current peer assessment activities is help for their learning achievement. | en_US |