Thesiss

Smoothing a probability distribution so that it generalizes well is a hard machine learning problem. It is particularly challenging when building a statistical language model with insufficient training data. We have developed a new smoothing algorithm (called AlphaRank) to overc…

A language model (LM) is a probability distribution over all possible word sequences. It is a vital component of many natural language processing tasks, such as automatic speech recognition, statistical machine translation, information retrieval and so on. The art of language mo…

We describe a statistical language model having components that are inspired by electrophysiological activities in the brain. These components correspond to important language-relevant event-related potentials measured using electroencephalography. We relate neural signals invol…

A language model is a vital component of automatic speech recognition systems. In recent years, advancements in neural network technologies have brought vast improvement in various machine learning tasks, including language modeling. However, compared to the conventional backoff…