Lexical Attraction Models (LAMs) were first introduced by Deniz Yuret in (Yuret 1998) to exemplify how an algorithm can learn word dependencies from raw text. His general thesis i...
Recent initiatives like the Million Book Project and Google Print Library Project have already archived several million books in digital format, and within a few years a significa...
Xiaoyue Wang, Lexiang Ye, Eamonn J. Keogh, Christi...
—We describe the design of an autonomous agent that can teach itself how to translate from a foreign language, by first assembling its own training set, then using it to improve...
Japanese case markers, which indicate the grammatical relation of the complement NP to the predicate, often pose challenges to the generation of Japanese text, be it done by a for...
This paper considers application of Deep Belief Nets (DBNs) to natural language call routing. DBNs have been successfully applied to a number of tasks, including image, audio and ...
Ruhi Sarikaya, Geoffrey E. Hinton, Bhuvana Ramabha...