International Journal of Engineering and Technology

As online social networks acquire larger user bases, they also become more interesting targets for spammers and promoters. Spam can take very different forms on social websites, especially in the form of videos and cannot always be detected by analyzing textual content. There are online video sharing systems that allow the users to post videos s response to any type of discussion topic. This feature encourages some of the users to post polluted content illegally as responses and there may be content promoters who try to promote them in the top listed search. Content pollution like spread advertise to generate sales, disseminate pornography, and compromise system reputation may threaten the trust of users on the system, thus weaken its success in promoting social interactions. As a solution for this problem, we classify the users as spammers, content promoters and legitimate users by building a test collection of real YouTube users using which we can provide a classification we use of content, individual and social attributes that help in characterizing each user class. For effective classification we use SVM-KNN which is an active learning approach. Our proposed approach poses a promising alternative to simply considering all users as legitimate or to randomly selecting users for manual inspection. In simple SVM training is very slow on whole dataset and not works very well on multiple classes. To overcome this problem and to provide efficient classification in fast manner we proposed new approach is SVM-KNN. Train a Support Vector Machine on K no of collections of nearest neighbours.