We will first build our own bigram language model, then create language models with a state-of-the-art tool which provides flexible smoothing options. We will make use of the language models in a specific task (predict the relative difficulty of a text). We will compare the power of different types of language models for the task and will examine some of the properties of the texts related to lexical use.
phaelishall / computational_linguistics_hw2 Goto Github PK
View Code? Open in Web Editor NEWWe will first build our own bigram language model, then create language models with a state-of-the-art tool which provides flexible smoothing options. We will make use of the language models in a specific task (predict the relative difficulty of a text). We will compare the power of different types of language models for the task and will examine some of the properties of the texts related to lexical use.