Word n-gram language model: Difference between revisions

Revision as of 17:19, 10 March 2023

Revision as of 15:41, 4 August 2023

Line 1:

{{DISPLAYTITLE:word ''n''-gram language model}}

AnA '''word n-gram language model''' is a [[language model]] that modelsgenerates sequencesprobabilities of words as a [[Markovseries process]].of Itwords, makesbased useon ofan the simplifying(over-simplified) assumption that the probability of the next word in a sequence depends only on a fixed size window of previous words. AIf bigram model considersonly one previous word is considered, it is called a trigrambigram model; considersif two words, anda intrigram general,model; anif ''n''-gram1 modelwords, considersan ''n''-1gram words of previous contextmodel.<ref name=jm/>

For example, a bigram language model modelswould assign the probabilityprobabilities of words in the sentence ''I saw the red house'' as:

<math display="block">P(\text{I, saw, the, red, house}) \approx P(\text{I}\mid\langle s\rangle) P(\text{saw}\mid \text{I}) P(\text{the}\mid\text{saw}) P(\text{red}\mid\text{the}) P(\text{house}\mid\text{red}) P(\langle /s\rangle\mid \text{house})</math>