Themes Updating Model Based on TIBG

Only considering freshness or total click number of a certain theme may lead to unreasonable themes updating. To improve the reasonableness of updating themes on homepage, this paper proposes a novel model based on theme interestingness of browser group (TIBG). TIBG can be used to calculate the theme's real-time popularity which tracks information related to theme browser's actual interest. Firstly, utilizing data mining methods, some data extracted from web log of AUST campus network are used to analyze the continuously changing trend of TIBG of every theme. Secondly, based on the achieved dynamic trend of every theme, we propose a model for updating multi-themes on homepage. According to this model, all TIBG of themes on a homepage can be calculated. Further, an integrated index reflecting the popularity of homepage can be obtained to measure the reasonableness of updating multi-themes. Experiment results show that our proposed model can implement reasonable updating of multi-themes and be effectively applied to the updating problems such as education resource, news theme on campus network.