The essay has to learn the concepts, terminology, techniques involving the method of detecting the user's comment habit or routine modeling user's comment, this is the basis to build the model the application personalization. Along with the development of the Web 2.0 generation, social networks, online news sites nowadays allow users can express a personal opinion about the content of the article/news online via the comment, share, love, opinion about the comments of other users ,... Based on the survey of the content of the comments to the news on the online news site VNExpress, analyzing the pros cons of the methods to build user models, the thesis proposes a solution that detects the user's comment habits based on the theme of probability. In particular, the system of enforcement of an algorithm evaluates ability to detect the user's habits by predicting the next comment may be happening. This approach has a lot of potential and has been proven through some statistics: initial results.
đang được dịch, vui lòng đợi..
