It is a well documented fact that, for human readers, familiar text is more legible than unfamiliar text. Current-generation computer vision systems also are able to exploit some ...
We propose two new tools to address the evolution of hyperlinked corpora. First, we define time graphs to extend the traditional notion of an evolving directed graph, capturing li...
Ravi Kumar, Jasmine Novak, Prabhakar Raghavan, And...
Advances in networking technology have enabled network engineers to use sampled data from routers to estimate network flow volumes and track them over time. However, low sampling ...
There are many algorithms to cluster sample data points based on nearness or a similarity measure. Often the implication is that points in different clusters come from different u...
Edward R. Dougherty, Junior Barrera, Marcel Brun, ...
We present an approach to classification of biomedical terms based on the information acquired automatically from the corpus of relevant literature. The learning phase consists of...