Scalable online incremental learning for web spam detection

Abstract

In this paper, we propose an online incremental learning framework for identifying web spam. The proposed work can incrementally update the learning model based on any newly arrived samples without recourse to the original data. The prototype of the framework has been evaluated with a real large scale web spam dataset. The results demonstrate the proposed online detector has high learning speed and accurate prediction rates for the web spam. © 2012 Springer-Verlag GmbH.