Educational Data Mining for Peer Assessment in Communities of Learners

摘要: In the last years, design and implementation of web-based education systems has grown exponentially, spurred by fact that neither students nor teachers are bound to a specific location this form computer-based is virtually independent any hardware platform. These accumulate large amount data: educational data mining learning analytics two much related fields research with aim using these improve process. chapter, authors investigate peer assessment setting in communities learners. Peer an effective didactic strategy, useful evaluate groups environments such as high schools universities where required answer open-ended questions increase their problem-solving skills. Furthermore, approach could become necessary contexts number be very as, for example, massive open online courses. Here author focus on automated support grading answers via evaluation-based approach, which mediated (partial) work teacher, produces (partial well) grading. The propose means methods, coming from data-mining field, Bayesian Networks K-Nearest Neighbours (K-NN), presenting some experimental results, our choices.