An unsupervised probabilistic learning framework for normalizing product records across different retailer Web sites is presented. Our framework decomposes the problem into two ta...
Bayesian methods are valuable, inter alia, whenever there is a need to extract information from data that is uncertain or subject to any kind of error or noise (including measurem...
In this paper, we describe a method for automatic creation of a knowledge source for text generation using information extraction over the Internet. We present a prototype system ...
We present an asymmetric watermarking method for copyright protection that uses different matrix operations to embed and extract a watermark. It allows for the public release of a...
Information on the World Wide Web becomes more and more important for our society. For blind people this is a chance to access more information for their everyday life. In this pap...