[−][src]Trait punkt::params::TrainerParameters
Configurable parameters for a trainer.
Associated Constants
const ABBREV_LOWER_BOUND: f64
Lower bound score for a token to be considered an abbreviation.
const ABBREV_UPPER_BOUND: f64
Upper bound score for a token to be considered an abbreviation.
const IGNORE_ABBREV_PENALTY: bool
Disables the abbreviation penalty which exponentially penalizes occurances of words without a trailing period.
const COLLOCATION_LOWER_BOUND: f64
Lower bound score for two tokens to be considered a collocation
const SENTENCE_STARTER_LOWER_BOUND: f64
Lower bound score for a token to be considered a sentence starter.
const INCLUDE_ALL_COLLOCATIONS: bool
Include all pairs where the first token ends with a period.
const INCLUDE_ABBREV_COLLOCATIONS: bool
Include all pairs where the first is an abbreviation. Overridden by
include_all_collocations
.
const COLLOCATION_FREQUENCY_LOWER_BOUND: f64
Minimum number of times a bigram appears in order to be considered a collocation.