Word Term Matrix

Question

I would love to create a Word matrix from some tweets, each word from the tweet has to be a new variable and be filled with 1 for only the words that correspond to that text in the tweet

x <- data.frame("Tweet" = c("hi all","I need help"), "N" = 1, "Reaction" = c("Happy", "Sad"), stringsAsFactors = FALSE)

I would love to paste the output but dont know how to do it honestly, sorry

enter image description here

See the `tm` package and the functions `DocumentTermMatrix` (or `TermDocumentMatrix`). — nicola, Jan 19 '18 at 16:10
also `quanteda`. the syntax is something like `df$Tweet %>% corpus() %>% dfm()` with `dplyr`. — amatsuo_net, Jan 19 '18 at 16:22

score 0 · Accepted Answer · answered Feb 18 '18 at 09:26

You could do it like this:

library(tm)

x <- data.frame("Tweet" = c("hi all","I need help"), "N" = 1, "Reaction" = c("Happy", "Sad"), stringsAsFactors = FALSE)

corp <- VCorpus(VectorSource(x$Tweet))
# adjust wordLengths, default is c(3, Inf)
dtm <- DocumentTermMatrix(corp, control = list(wordLengths = c(1, Inf)))
data.frame(Tweet = x$Tweet, as.matrix(dtm), Reaction = x$Reaction)

            Tweet all help hi i need Reaction
1      hi all   1    0  1 0    0    Happy
2 I need help   0    1  0 1    1      Sad

Word Term Matrix

1 Answers1