Text.Language.Classifier behaviour (Text v0.2.0) View Source
A behaviour definition module for language classifiers.
A langauge classifier correlates supplied natural language text against a vocabulary and returns a score indicating how closely the supplied text matches the vocabulary.
Link to this section Summary
Types
A list of 2-tuples of the form {language, number}
A classifier is a module that implements the Text.Language.Classifier behaviour.
A list mapping an n-gram as a charlist to a Text.Ngram.Frequency struct
Callbacks
Sorts the classifier scores from all languages in order or correlation.
Returns the classifier score for one language.
Link to this section Types
Specs
frequency_list() :: [Text.frequency_tuple(), ...]
A list of 2-tuples of the form {language, number}
Specs
t() :: module()
A classifier is a module that implements the Text.Language.Classifier behaviour.
Specs
text_ngrams() :: %{required(charlist()) => Text.Ngram.Frequency.t()}
A list mapping an n-gram as a charlist to a Text.Ngram.Frequency struct
Link to this section Callbacks
Specs
order_scores(frequency_list()) :: frequency_list()
Sorts the classifier scores from all languages in order or correlation.
Specs
score_one_language(Text.language(), text_ngrams(), Text.vocabulary()) :: frequency_list()
Returns the classifier score for one language.
A classifier correlates how closely a supplied string (encoded into n-grams) matches against a given langauge profile implemented as a vocabulary.
See Text.Language.Classifier.NaiveBayesian
for an example.