Research On Recommendation System Of Heterogeneous Peer In MCSCL

With the development and popularization of mobile technology, mCSCL has become a newhotspot in education technology.mCSCL supports the interactive between peers and developscollaborative learning through mobile technology. Appropriate grouping of collaborative learning isthe premise to carry out the activity smoothly.The group members that always be called peer andgrouping strategy are the key issue in collaborative grouping.In the CL field, heterogeneous groupstrategy is the consensus of the researchers, it also has advantages.So,how to recommend appropriateheterogeneous peer is one of the key problem in mCSCL, and it is also problem that we want to solve.However,searching peers for learner and then formatting a collaborative group by humanaccording to a personality characteristic of learner is the main grouping strategy;there is lacking aneffective recommend model for peer grouping.There are existing three problem in peer recommending:（1）The standard of formatting a mCSCL peer is too onefold and haven’t comprehensively considerthe personality characteristic of learner and the particularity of mCSCL, so that there isn’t existing apeer model for peer formatting.（2）Emphasised artificial grouping too much,there isn’t existing aappropriate heterogeneous peer recommend algorithm.（3）There isn’t existing a peer recommendingsystem that provide choice for collaborative grouping in mCSCL.To solve the problems,the paper put forwards a research on recommend system ofheterogeneous peer in mCSCL.The work mainly includes the following three aspects:Firstly, the paper analysises the key factor that should be considered when we select our peers inmCSCL.At the same time,we build the peer model that is suitable for mCSCL and finish formalizationand digital of the peer model.Secondly, according to the peer model in mCSCL,we design the recommend algorithm forheterogeneous peer based on K-means.And then,the paper providers the recommend model forheterogeneous peer in mCSCL. Lastly,on basis of the above research,this paper design and develop the recommend system forheterogeneous peer in mCSCL.This paper develop experiments and analysis experiment data.Theresults show that the system can help search the heterogeneous peer for learner and format the group.The search can improve the organization form and learning efficiency of mCSCL, enrich researchon peer and group in mCSCL.The Mpr can promote the search of heterogeneous peer for learner inmCSCL.At the same time,I hope that the Mpr can provide some choices for the collaborationgrouping.