This paper describes the result of performance evaluation of two kinds of MapReduce applications running in the FutureGrid: a data intensive application and a computation intensive...
— The development of a software product in an organization involves various groups of stakeholders who may prioritize the qualities of the product differently. This paper present...
In this paper we present clustering method is very sensitive to the initial center values ,requirements on the data set too high, and cannot handle noisy data the proposal method ...
We present a simple semi-supervised relation extraction system with large-scale word clustering. We focus on systematically exploring the effectiveness of different cluster-based ...
We propose a method to enhance multi-stream Gabor and MFCC features using data-driven hierarchical phoneme clusters to yield more discriminating posteriors. We take into account d...