stylest_predict: Predict the most likely speaker of a text

In stylest: Estimating Speaker Style Distinctiveness

Description

Use a fitted stylest_model to predict the most likely speaker of a text. This function may be used on in-sample or out-of-sample texts.

Usage

stylest_predict(model, text, prior = NULL)

Arguments

model	stylest_model object
text	Text vector. May be a corpus_frame object
prior	Prior probability, defaults to NULL

Value

stylest_predict object containing: model the fitted stylest_model object used in prediction, predicted the predicted speaker, log_probs matrix of log probabilities, log_prior matrix of log prior probabilities

Examples

data(novels_excerpts)
speaker_mod <- stylest_fit(novels_excerpts$text, novels_excerpts$author)
stylest_predict(speaker_mod, "This is an example text, who wrote it?")
  

Example output

$model
A S3 stylest_model object containing: 3 unique authors and 1043 unique terms.
$predicted
[1] Austen, Jane
Levels: Austen, Jane Eliot, George Gaskell, Elizabeth Cleghorn

$log_probs
1 x 3 Matrix of class "dgeMatrix"
     Austen, Jane Eliot, George Gaskell, Elizabeth Cleghorn
[1,]   -0.3211882     -1.405527                   -3.524201

$log_prior
               Austen, Jane               Eliot, George 
                  -1.098612                   -1.098612 
Gaskell, Elizabeth Cleghorn 
                  -1.098612 

attr(,"class")
[1] "stylest_predict"

stylest documentation built on March 5, 2021, 1:05 a.m.