An application of SVM: Blog templates recommendation system

This paper demonstrates an application, blog templates recommendation system, applied with Support Vector Machine (SVM). In recent years, the population of Blog users keeps growing rapidly. This study uses SVM to be a training method for creating a recommendation module. When the new user´s data have been got into the modules, it will produce a suitable CSS template. Users can use the generated template file to change the CSS templates and apply to the Blog page in order to achieve the best results. We gather the users´ attributes via simple and intuitive steps of the web page operations from themselves. The logs of web page will be sent back-end to database and record the user´s preferences and settings for feedback schema immediately. After training all of the user data, system will provide the best template for user. It can save time for users to the selection and design CSS. Also, we use QUIS to do post-test questionnaire, the result is satisfactory. Finally, the discussion in this paper proposed a new methodology for classification of SVM application.