Identifying synonymous rows of a text column in a dataframe using R

Question

Suppose ABC is a dataframe as given below:

ABC <- data.frame(Column1 = c(1.222, 3.445, 5.621, 8.501, 9.302), 
                  Column2 = c(654231, 12347, -2365, 90000, 12897), 
                  Column3 = c('A1', 'B2', 'E3', 'C1', 'F5'), 
                  Column4 = c('I bought it', 'The flower has a beautiful fragrance', 'It was bought by me', 'I have bought it', 'The flower smells good'), 
                  Column5 = c('Good', 'Bad', 'Ok', 'Moderate', 'Perfect'))

My intention is to find synonymous strings in Column4. In this case, I bought it, It was bought by me and I have bought it are synonymous or similar strings and The flower has a beautiful fragrance and The flower smells good convey similar meaning.

I tried the approach of IVR in the following thread and got stuck: Find similar texts based on paraphrase detection

When I run the HLS.Extract code chunk, I get the following error message:

Error in strsplit(PlainTextDocument(synonyms(word)), ",") : non-character Argument

Using as.character didn't resolve the problem either:

Syns = function(word){  
    word <- as.character(word) ###
    wl    =   gsub("(.*[[:space:]].*)","",      
                   gsub("^c\\(|[[:punct:]]+|^[[:space:]]+|[[:space:]]+$","",  
                        unlist(strsplit(PlainTextDocument(synonyms(word)),","))))
    wl = wl[wl!=""] 
    return(wl)     
  }

What is going wrong?
Is there a better way to code it using R, and create a new column additionally which has, say a number 1 as an entry for the first synonymous strings and 2 as the entry for the next set of synonymous strings?
Does it work with German text?

score 0 · Answer 1 · answered Jan 25 '21 at 12:17

The issue was resolved by setting PlainTextDocument(synonyms(word)) as character, as shown below:

Syns = function(word){ 
    wl    =   gsub("(.*[[:space:]].*)","",      
                   gsub("^c\\(|[[:punct:]]+|^[[:space:]]+|[[:space:]]+$","",  
                        unlist(strsplit(as.character(PlainTextDocument(synonyms(word))),",")))) 
    wl = wl[wl!=""] 
    return(wl)     
  }

Identifying synonymous rows of a text column in a dataframe using R

1 Answers1