This paper explores two classes of model adaptation methods for Web search ranking: Model Interpolation and error-driven learning approaches based on a boosting algorithm. The res...
Jianfeng Gao, Qiang Wu, Chris Burges, Krysta Marie...
An appreciation of the roles of genre and task is important in understanding how people browse the Web. Genre is characterized by content and form and is intimately linked to the ...
Carolyn R. Watters, Michael A. Shepherd, Forbes J....
Detection of web attacks is an important issue in current defense-in-depth security framework. In this paper, we propose a novel general framework for adaptive and online detectio...
Wei Wang 0012, Florent Masseglia, Thomas Guyet, Re...
This paper presents a real-world application for assisting medical diagnosis and drug prescription, which relies on the exclusive use of machine learning techniques. We have autom...
This paper1 presents an empirical approach to mining parallel corpora. Conventional approaches use a readily available collection of comparable, nonparallel corpora to extract par...