Labeling text data is quite time-consuming but essential for automatic text classification. Especially, manually creating multiple labels for each document may become impractical ...
All Netflix Prize algorithms proposed so far are prohibitively costly for large-scale production systems. In this paper, we describe an efficient dataflow implementation of a coll...
Srivatsava Daruru, Nena M. Marin, Matt Walker, Joy...
Routing is one of the most challenging open problems in disruption-tolerant networks (DTNs) because of the shortlived wireless connectivity environment. To deal with this issue, r...
Information diffusion, viral marketing, and collective classification all attempt to model and exploit the relationships in a network to make inferences about the labels of nodes....
In some applications such as filling in a customer information form on the web, some missing values may not be explicitly represented as such, but instead appear as potentially va...