This paper presents a new approach to language model construction, learning a language model not from text, but directly from continuous speech. A phoneme lattice is created using acoustic model scores, and Bayesian techniques are used to robustly learn a language model from this noisy input. A novel sampling technique is devised that allows for the integrated learning of word boundaries and an...

the main purpose of this paper has been to predict english students’ academic achievement on the basis of achievement goals, motivational –usefulness beliefs and the kind of mobile usages. the population of the study consisted of english language students in the faculty of foreign language at isfahan university from both undergraduate and graduate levels in 2011-2012. by applying a random sampl...

Softmax is the most commonly used output function for multiclass problems and is widely used in areas such as vision, natural language processing, and recommendation. A softmax model has linear costs in the number of classes which makes it too expensive for many real-world problems. A common approach to speed up training involves sampling only some of the classes at each training step. It is kn...

Oscillatory entrainment to the speech signal is important for language processing, but has not yet been studied in developmental disorders of language. Developmental dyslexia, a difficulty in acquiring efficient reading skills linked to difficulties with phonology (the sound structure of language), has been associated with behavioural entrainment deficits. It has been proposed that the phonolog...

OBJECTIVE Children with developmental language disorders have been shown to be impaired not only in language processing (including syntax), but also in rhythm and meter perception. Our study tested the influence of external rhythmic auditory stimulation (i.e., musical rhythm) on syntax processing in children with specific language impairment (SLI; Experiment 1A) and dyslexia (Experiment 1B). ...

As probabilistic computations play an increasing role in solving various problems, researchers have designed probabilistic languages to facilitate their modeling. Most of the existing probabilistic languages, however, focus only on discrete distributions, and there has been little effort to develop probabilistic languages whose expressive power is beyond discrete distributions. This dissertatio...