sentiment_datumbox.r

# load packages
library(twitteR)
library(RCurl)
library(RJSONIO)
library(stringr)



getSentiment <- function (text, key){
  
  
  
  text <- URLencode(text);
  
  #save all the spaces, then get rid of the weird characters that break the API, then convert back the URL-encoded spaces.
  text <- str_replace_all(text, "%20", " ");
  text <- str_replace_all(text, "%\\d\\d", "");
  text <- str_replace_all(text, " ", "%20");
  
  
  if (str_length(text) > 360){
    text <- substr(text, 0, 359);
  }
  ##########################################
  
  data <- getURL(paste("http://api.datumbox.com/1.0/TwitterSentimentAnalysis.json?api_key=", key, "&text=",text, sep=""))
  
js <- fromJSON(data, asText=TRUE);
 
  # get mood probability
  sentiment = js$output$result

###################################

  data <- getURL(paste("http://api.datumbox.com/1.0/SubjectivityAnalysis.json?api_key=", key, "&text=",text, sep=""))
  
js <- fromJSON(data, asText=TRUE);
 
  # get mood probability
  subject = js$output$result

##################################

data <- getURL(paste("http://api.datumbox.com/1.0/TopicClassification.json?api_key=", key, "&text=",text, sep=""))
  
js <- fromJSON(data, asText=TRUE);
 
  # get mood probability
  topic = js$output$result

##################################
data <- getURL(paste("http://api.datumbox.com/1.0/GenderDetection.json?api_key=", key, "&text=",text, sep=""))
  
js <- fromJSON(data, asText=TRUE);
 
  # get mood probability
  gender = js$output$result

return(list(sentiment=sentiment,subject=subject,topic=topic,gender=gender))
  }
      
 

#################
clean.text <- function(some_txt)
{
   some_txt = gsub("(RT|via)((?:\\b\\W*@\\w+)+)", "", some_txt)
   some_txt = gsub("@\\w+", "", some_txt)
   some_txt = gsub("[[:punct:]]", "", some_txt)
   some_txt = gsub("[[:digit:]]", "", some_txt)
   some_txt = gsub("http\\w+", "", some_txt)
   some_txt = gsub("[ \t]{2,}", "", some_txt)
   some_txt = gsub("^\\s+|\\s+$", "", some_txt)

   # define "tolower error handling" function
   try.tolower = function(x)
   {
      y = NA
      try_error = tryCatch(tolower(x), error=function(e) e)
      if (!inherits(try_error, "error"))
         y = tolower(x)
      return(y)
   }

   some_txt = sapply(some_txt, try.tolower)
   some_txt = some_txt[some_txt != ""]
   names(some_txt) = NULL
   return(some_txt)
}


# harvest tweets
tweets = searchTwitter("iPhone", n=200, lang="en")


# get text
tweet_txt = sapply(tweets, function(x) x$getText())


# clean text
tweet_clean = clean.text(tweet_txt)

#####################################
# how many tweets
tweet_num = length(tweet_clean)

# data frame (text, sentiment, score)
tweet_df = data.frame(text=tweet_clean, sentiment=rep("", tweet_num),
   subject=1:tweet_num, topic=1:tweet_num, gender=1:tweet_num, stringsAsFactors=FALSE)

# apply function getSentiment
sentiment = rep(0, tweet_num)
for (i in 1:tweet_num)
{
   tmp = getSentiment(tweet_clean[i], "API_KEY")
   
	tweet_df$sentiment[i] = tmp$sentiment
  
	tweet_df$subject[i] = tmp$subject
	tweet_df$topic[i] = tmp$topic
	tweet_df$gender[i] = tmp$gender
}