Research On Curriculum Recommendation Algorithm Based On Restricted Boltzmann Machine Collaborative Filtering And Hadoop-Mahout

Deep learning has been developing rapidly in many fields since it was put forward.Deep learning application research is mainly related to the field of computer and educational technology.There are different concepts research objects and roles in the two fields.In the field of educational technology,it is mainly to promote students' deep learning and improve the quality of education in discipline teaching.Meanwhile,in the field of educational technology,we also use the concept of deep learning in computer field to carry out research and development of education and teaching.Recommendation algorithm as the main application of collaborative filtering algorithm in Internet application plays a leading role,convolutional neural network.RBM and depth of confidence plays an important part in the research and application of network of neural network model in deep learning.which restricted Boltzmann machine in the collaborative filtering recommendation algorithm is the most widely used in.In this paper,the application of the limited Boltzmann machine in the recommendation algorithm of collaborative filtering is studied,The application of the MOOC education curriculum recommendation is put forward.At present,the online education of MOOC has been developing and the number of online learners is increasing.In view of the fact that MOOC online learners are not easy to find suitable courses in massive curriculum information.Blind selection leads to low learning efficiency and poor learning effect.In this paper,a restricted Boltzmann RBM recommendation algorithm is proposed.In this algorithm,IRS curriculum evaluation index system is put forward,which carries out a unified standard evaluation of online courses.Combines and improves the machine learning framework Apache-Mahout collaborative filtering recommendation algorithm recommending personalized courses for online learners.Based on the Hadoop distributed cloud computing platform.The data of a large number of learners are analyzed and the course recommendation is made.Experimental results show that the recommendation algorithm proposed in this paper is effective and applicable to distributed cloud computing environment.And verifies the efficiency of parallel computing in distributed environment.