Looks up specified ngramMap for the given token and returns a map of the frequency for each dependency label for the given token word and POS, normalized over the total frequency for all possible dependency labels.
Looks up specified ngramMap for the given token and returns a map of the frequency for each dependency label for the given token word and POS, normalized over the total frequency for all possible dependency labels.
the token to look up
the table mapping a word to the sequence of NgramInfos, as obtained from the GoogleNGram class object
the frequency cutoff that was used to construct the map. This is used here to shift the scale of the frequencies to start from the cutoff point instead of 1.
Method to bucketize a given normalized frequency for feature generation.
Looks up specified ngramMap for the given word and returns a map of the frequency for each POS tag for the given word, normalized over the total frequency for all possible POS tags.
Looks up specified ngramMap for the given word and returns a map of the frequency for each POS tag for the given word, normalized over the total frequency for all possible POS tags.
the word to look up
the table mapping a word to the sequence of NgramInfos, as obtained from the GoogleNGram class object
the frequency cutoff that was used to construct the map. This is used here to shift the scale of the frequencies to start from the cutoff point instead of 1.
Object encapsulating some functionality specific to unigrams. Used wherever features need to be constructed based on unigrams (Google Ngram Nodes).