Abstract
With the popularity of Facebook, it has become the newest platform of spamming distributor. Our target is focusing on Facebook group pages, which is occupied with spam messages. With the huge amount of spam, user might feel disturbed, and even confront personal data leak. Therefore, it is necessary to solve this problem. We implement SVM algorithms to support our text filtering mechanism. And according to the group's features to evaluate each one. Lastly, based on the results, to send the appropriate advices to the users. We hope each user could have a better experience on Facebook.