Decision-tree algorithms are known to be unstable: small variations in the training set can result in different trees and different predictions for the same validation examples. B...
Event extraction is a particularly challenging type of information extraction (IE). Most current event extraction systems rely on local information at the phrase or sentence level...
Conditional random fields (CRF) are widely used for predicting output variables that have some internal structure. Most of the CRF research has been done on structured classificati...
Social recommendation, that an individual recommends an item to another, has gained popularity and success in web applications such as online sharing and shopping services. It is l...
This paper presents a novel system that employs an adaptive neural network for the no-reference assessment of perceived quality of JPEG/JPEG2000 coded images. The adaptive neural ...