Research On Learner Preference Model In M-Learning

Mobile learning is a new stage in the development of education after the D-Learning, and E-Learning approach to learning. Learners can achieve the freedom to learn at any time, any place, get rid of the limit of time and space.In general, the Web learning services website contains many knowledge modules, informative, rich in content. The learners want through mobile terminals to quickly find knowledge module or theme of the web site that they are interested in, it is not easy. This is because the phone screen is small, you have to go through the scroll off the screen, search and other operations in order to achieve preference information or demand information.If the contents of the Web learning services website can change along with the different mobile users, according to different learners to meet the needs of learning content, learners can save a large amount of operating time, at the same time increase the learners’ reliance on the Web learning services website. Therefore provide a personalized learning service in a mobile environment for different learners is very urgent. The user preferences modeling is the foundation and core of personalized recommendation service.The main work of this paper is as follows:(1) Introduce user preference modeling, Web data mining technology, ontology and other related technical and theoretical knowledge, and these provide a theoretical basis and technical support for the follow-up work unfolds.(2) The establishment of learner preferences model. We build learner preference model introduced Ontology, improve the ability to describe learner preferences and semantic skills.we use the Ontology design tools protege and the W3C’s ontology language OWL builds the model.(3) Get learner preference information. Get learner preference information that is a basic and important work in the user preferences model. This paper mining learner preference data from Web logs. Before mining algorithm, the first pre-processing of log data, and then extract content preferences and media preference information from the user log records. Get learner content preference information, use DBSCAN clustering algorithm, and the algorithm to do a modified accordingly. Finally, the experimental verification of the contents of clustering results is feasible and effective.The main work of this paper is to establish an appropriate mathematical model to describe learner behavior, build specification learners description file, and use of data mining algorithms to extract the learner preference characteristics, provide strong support for the follow-up to provide personalized learning services to the learners.