We present two methods for estimating replacement probabilities without using parallel corpora. The first method proposed exploits the possible translation probabilities latent in ...
Background: Drug-drug interactions are frequently reported in the increasing amount of biomedical literature. Information Extraction (IE) techniques have been devised as a useful ...
Abstract. This research explores a new method for Semantic Web service matchmaking based on iSPARQL strategies, which enables to query the Semantic Web with techniques from traditi...
The problem of transfer learning, where information gained in one learning task is used to improve performance in another related task, is an important new area of research. While...
We present a large-scale analysis of the content of weblogs dating back to the release of the Blogger program in 1999. Over one million blogs were analyzed from their conception t...