The emergence of pervasive networked data sources, such as web services, sensors, and mobile devices, enables context-sensitive, mobile applications. We have developed a programmi...
Norman H. Cohen, Hui Lei, Paul Castro, John S. Dav...
File I/O data is interpreted by high performance parallel/distributed applications mostly as a sequence of arbitrary bits. This leads to the situation where data is ’volatile’...
This paper studies the effects of training data on binary text classification and postulates that negative training data is not needed and may even be harmful for the task. Tradit...
Abstract. Our research group has analyzed many industrial, frameworkbased applications. In these applications, simple functionality often requires excessive runtime activity. It is...
We present a unified model of what was traditionally viewed as two separate tasks: data association and intensity tracking of multiple topics over time. In the data association pa...