The Automatic Evaluation Strategies and Methods of Multimedia Work Assignments

Grading students\´ assignments of multimedia work is the key step in the "multimedia making" courses. Now most multimedia work assignments (e.g. Photoshop images, Flash animations) are just evaluated by the teachers manually. This evaluation method is not only laborious, tedious and time-consuming, but also difficult to avoid manual errors. Hence, it is necessary to automate the grading procedure for the purpose of reducing workload and subjectivity. This paper introduces a set of methods for auto-evaluating Photoshop images and Flash animations, which includes multi-level indicator systems and fuzzy matching evaluation strategies for multimedia assignments. Moreover, the Automatic Evaluation System of Multimedia Work Assignments (AESMWA) has been successfully employed in the high school graduating test of Zhejiang Province, China. Application results show its excellent performance in both efficiency and accuracy.