With this App, users can build a wordcloud for any Platonic dialogue and to check the frequency of words in it. The lemmatized text was extracted from the Diorisis Ancient Greek Corpus and parsed in R.

Try it out! https://locusclassicus.shinyapps.io/myshinyapp/

The function which runs under the hood transforms dialogues selected by users into a TermDocumentMatrix, deletes stopwords (i.e., most frequent ‘function’ words such as “and”, “or”, etc.), and builds a cloud. Here is the code for it.

make.wordcloud <- function(x) {
  my_text <- my_corpus[[x]]
  text.sw <- my_text %in% sw
  text.clean <- my_text[!text.sw]
  doc <- Corpus(VectorSource(text.clean))
  dtm <- TermDocumentMatrix(doc) 
  matrix <- as.matrix(dtm)
  words <- sort(rowSums(matrix),decreasing=TRUE) 
  df <- data.frame(word = names(words),freq=words)
  wordcloud2(data=df, size=1, color='random-dark', shape = "circle")
}

Word frequencies can be a powerful tool not only for authorship attribution, but also for topic modelling. And wordclouds vizualize them is a very informative way. We could further study frequencies themselves: thus, the interrogative “who/what” is strongly correlated to the average number of replies per page (see next slide).

library(ggplot2)
ggplot(df, aes(x=ratio, y=τίς, color=type)) +
  geom_point()