Analysis Of The Consumer Learning Preference Based On Clustering Method

Information-based learning environment,which cause widespread attention in recent years, as well as MOOC, online courses, learning space mode applications,which continue to emerge, and information-based learning environment is the main advantage of features is to achieve user-oriented personalized learning services. To do this, we need a set of information-based learning environments intelligent analysis mechanism called preference analysis mechanism,which lay the foundation for individualized learning services. Therefore learning preference analysis has an significent application value in information learning environment.In the application of information technology conditions, server will record user logs operated by the learners,which are able to reflect the user's learning preferences. Through analysising and carding,I select the URL requested by the user, the user visits, resource's frequency used by users, user access time, content type, etc,which can be used to get the user's learning preferences.In the analysis of indicators, preprocessing all original data when extract data and eliminating'noisy point' by identifying elements which is irrelevant to preference. Using a regular expression matching, extract various sources of data to be processed preference and storing like text files or database tables. In the data analysis, the use of density-based clustering analysis method, adapted to the learning preferences that using density as explicit feature set of data processing,as well as those can not be pre-biased classification of learning preferences analysis. In addition, due to differences in individual characteristics of the user, cluster analysis parameter values can not be consistent, so when selecting parameter analysis method,using the dynamic rules to adapting the current user behavior. After the data analysis is completed, results are expressed using XML language,playing the advantages which may be extended semantics and machine-readable representing the uses learning preference characteristics.