In an embedded system, it is common to have several memory areas with different properties, such as access time and size. An access to a specific memory area is usually restricted...
For many machine learning solutions to complex applications, there are significant performance advantages to decomposing the overall task into several simpler sequential stages, c...
We propose a hybrid generative/discriminative framework for semantic parsing which combines the hidden vector state (HVS) model and the hidden Markov support vector machines (HMSV...
A common limitation of many retrieval models, including the recently proposed axiomatic approaches, is that retrieval scores are solely based on exact (i.e., syntactic) matching o...
: Educational standards alignment, which matches similar or equivalent concepts of educational standards, is a necessary task for educational resource retrieval. In order to automa...