CANDIDATESCORE {MoLE}

Score candidate expressions

Description

Provides each candidate expression for some meaning or function with a score in which (depending on the model settings) semantic match, lexeme activation, (relative) frequency of use, recency, collostruction frequency, semantic weight, and/or economy of expression are taken into consideration.

Usage

CANDIDATESCORE(lexicon, type = "referringExpression")

Arguments

lexicon

lexicon with candidate expresions

type

Type of function for which an expression has to be found (referringExpression, nounMarker, verbMarker, or pronoun).

Details

Collostruction frequencies are determined differently for different type of functions. The lighter, the better; recency starts with 0.

Value

Vector of scores, corresponding to the entries evaluated.

Note

Match and collostruction frequency are calculated separately before CANDIDATESCORE can apply. In the example below, the latter is randomly set for illustration purposes.

Author(s)

Sander Lestrade

See Also

SELECTVERB, SELECTACTOR, SELECTUNDERGOER, REFCHECK, TOPICCOPY, GENERALIZE, CHECKSUCCESS

Examples

FOUND()
lexicon=head(population[[1]]$nouns)
lexicon$match=VMATCH(lexicon[1,1:9], lexicon)
lexicon$collostruction=sample(100, nrow(lexicon))	
lexicon$score=CANDIDATESCORE(lexicon)