|
gptkbp:instanceOf
|
Language Model
|
|
gptkbp:alternativeTo
|
gptkb:Unigram_Language_Model
Neural Language Model
Trigram Language Model
|
|
gptkbp:application
|
gptkb:Machine_Translation
gptkb:Speech_Recognition
Text Prediction
|
|
gptkbp:assumes
|
gptkb:Markov_property
each word depends only on the previous word
|
|
gptkbp:canBeSmoothedWith
|
Good-Turing Smoothing
Kneser-Ney Smoothing
Laplace Smoothing
|
|
gptkbp:developedBy
|
20th century
|
|
gptkbp:estimatedCost
|
probability of word sequences
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Bigram Language Model
|
|
gptkbp:input
|
sequence of words
|
|
gptkbp:limitation
|
data sparsity
limited context
|
|
gptkbp:output
|
probability distribution over next word
probability of sequence
|
|
gptkbp:relatedTo
|
gptkb:Hidden_Markov_Model
Statistical Language Model
|
|
gptkbp:requires
|
training corpus
|
|
gptkbp:type
|
gptkb:N-gram_Language_Model
|
|
gptkbp:usedBy
|
search engines
text editors
|
|
gptkbp:usedIn
|
gptkb:Natural_Language_Processing
|
|
gptkbp:uses
|
bigrams
|
|
gptkbp:bfsParent
|
gptkb:Unigram_Language_Model
|
|
gptkbp:bfsLayer
|
8
|