This paper introduces the application of data mining methods to the analysis and prediction of the typhoon. The testbed for this research is the typhoon image collection that we e...
We present a class of richly structured, undirected hidden variable models suitable for simultaneously modeling text along with other attributes encoded in different modalities. O...
In this paper, we propose a methodology to predict the popularity of online contents. More precisely, rather than trying to infer the popularity of a content itself, we infer the l...
Abstract— Sensor networks have evolved to a powerful infrastructure component for event monitoring in many application scenarios. In addition to simple filter and aggregation op...
Daniel Klan, Katja Hose, Marcel Karnstedt, Kai-Uwe...
This paper presents a new prediction model for predicting when an online customer leaves a current page and which next Web page the customer will visit. The model can forecast the ...