How can I create a tf-idf matrix with character n-gram features?

Question

How can I use the text2vec package to create a tdf-idf matrix with character n-gram features?

Answer 1

How about:

library(text2vec)
data("movie_review")
it = itoken(movie_review$review, tolower, char_tokenizer)
v = create_vocabulary(it, ngram = c(3, 3), sep_ngram = "_")
dtm = create_dtm(it, vectorizer = vocab_vectorizer(v))

PS in future please try to provide some reproducible example of what did you try to solve your problem.