Twitter Sentiment Analysis in R with Keyword #Modi

Few months back my colleagues were discussing about this kind of analysis. We were given assignment as well but didn’t take serious at that time. Yesterday I was Googling some article and found this interesting analysis and tried myself. As I am a follower of Modi and he has million of followers on Twitter, I thought of doing ‘Sentiment Analysis’ on #MODI keyword. This is interesting as Narendra Modi (often called as Modi) is the current Prime Minister of India and he is very active in social media. But India is a democratic country and everyone has right to speak so let’s see how people are reacting to Modi.

Before starting the program let me put something forward:

1. The idea is copied from Google and code is written from the help of various sources
2. You must have the below mentioned libraries
3. You need to have at least R 2.15.1 or above to run the following code
4. You need to have own Twitter account
5. You can’t run the entire code in single shot, have to run line by line

You can download the list of positive and negative words from below links:

Negative Words
Positive Words

Now you need to create a Twitter application:

1. Go to https://apps.twitter.com/
2. Create a new Twitter app
3. Fill the boxes accordingly and make it private
4. After doing the formality you will this kind of page
5. If you click on “Manage keys and access token”, you will get a ‘Customer Secret’
6. Save this information. We will use it later.

####################### TWITTER SENTIMENT ANALYSIS ###################
############# AUTHOR: DEEPESH SINGH
############# PURPOSE: TWITTER SENTIMENT ANALYSIS WITH KEYWORD #MODI
############# DATE: 12/20/2015


########################### MODULE 1 ################################
############# HANDSHAKING TWITTER AND GETTING THE DATA ##############

## SETTING WORKING DIRECTORY
path <- "E:/R Programs/Twitter Sentiment Analysis"
setwd(path)

## CALLING ALL REQUIRED LIBRARY
library(twitteR)     #twitter interface API
library(ROAuth)      #for authentication
library(plyr)        #to break big problem into smaller
library(stringr)     #make easier to work with strings
library(ggplot2)     #implementation of the grammar of graphics
library(RColorBrewer)#provide palettes for drawing
library(tm)          #tools for text mining
library(wordcloud)   #used to create word cloud
library(RCpp)        #R and C++ integration
library(RCurl)       #HTTP/FTP client interface for R


##
## Bundle of CA Root Certificates
##
## Certificate data from Mozilla as of: Wed Oct 28 04:12:04 2015
##
## This is a bundle of X.509 certificates of public Certificate Authorities
## (CA). These were automatically extracted from Mozilla's root certificates
## file (certdata.txt).  This file can be found in the mozilla source tree:
## http://hg.mozilla.org/releases/mozilla-release/raw-file/default/security/nss/lib/ckfw/builtins/certdata.txt
##
## It contains the certificates in PEM format and therefore
## can be directly used with curl / libcurl / php_curl, or with
## an Apache+mod_ssl webserver for SSL client authentication.
## Just configure this file as the SSLCACertificateFile.
##
## Conversion done with mk-ca-bundle.pl version 1.25.
## SHA1: 6d7d2f0a4fae587e7431be191a081ac1257d300a
##
## DOWNLOAD FILE
download.file(url="http://curl.haxx.se/ca/cacert.pem", destfile = "cacert.pm")

## NOW WE WILL ACCESS THE TWITTER API WHICH WE HAVE JUST CREATED
## ALL DETAILS CAN BE FOUND FROM THE TWITTER APPLICATION PAGE
requestURL <- "https://api.twitter.com/oauth/request_token"
accessURL <- "https://api.twitter.com/oauth/access_token"
authURL <- "https://api.twitter.com/oauth/authorize"
consumerKey <- "LMfRAhWF##############2a"
consumerSecret <- "06rghyRRxwLE4################HW2h63iPISeQkxCB4rf"
Cred <- OAuthFactory$new (consumerKey = consumerKey,
                          consumerSecret = consumerSecret,
                          requestURL = requestURL,
                          accessURL = accessURL,
                          authURL = authURL)

Cred$handshake(cainfo = system.file("CurlSSL", "cacert.pem", package =  ('RCurl')))

## AFTER RUNNING THIS LINE YOU WILL GET A URL WHICH YOU NEED TO PASTE TO YOUR WEB BROWSER TO GET A KEY
## IT WILL LOOK LIKE THIS MESSAGE
## To enable the connection, please direct your web browser to: 
## https://api.twitter.com/oauth/authorize?oauth_token=iTftDgAAAAAAjWK9AAABUcCT6T8
## When complete, record the PIN given to you and provide it here: 7047697

## NOW WE WILL SAVE THIS ENVIRONMENT
save(Cred, file = "Twitter_Auth.RData")
load("Twitter_Auth.RData")
registerTwitterOAuth(Cred)


## THE CONSOLE OUTPUT WILL BE TRUE THAT MEANS WE HAVE GOT SUCCESSFUL HANDSHAKE

## NOW WE WILL TRY TO FETCH TWITS FROM TWITTER USING THE HASH TAG KEYWORD
download.file(url="http://curl.haxx.se/ca/cacert.pem", destfile="cacert.pem")
## HERE I AM EXTRACTING TOP 500 TWITS ON THIS TOPIC/KEYWORD
modi.list <- searchTwitter('#Modi', n = 500, cainfo = "cacert.pem")
modi.df <- twListToDF(modi.list)
write.csv(modi.df, "ModiTwits.csv",row.names = F)


########################### MODULE 2 ################################
############# SENTIMENT ANALYSIS WITH THE ABOVE DATA  ##############


## SENTIMENT SCORE = POSITIVE SCORE - NEGATIVE SCORE
# function score.sentiment
score.sentiment = function(sentences, pos.words, neg.words, .progress='none')
{
  # Parameters
  # sentences: vector of text to score
  # pos.words: vector of words of postive sentiment
  # neg.words: vector of words of negative sentiment
  # .progress: passed to laply() to control of progress bar
  
  # create simple array of scores with laply
  scores = laply(sentences,
                 function(sentence, pos.words, neg.words)
                 {
                   # remove punctuation
                   sentence = gsub("[[:punct:]]", "", sentence)
                   # remove control characters
                   sentence = gsub("[[:cntrl:]]", "", sentence)
                   # remove digits?
                   sentence = gsub('\\d+', '', sentence)
                   
                   # define error handling function when trying tolower
                   tryTolower = function(x)
                   {
                     # create missing value
                     y = NA
                     # tryCatch error
                     try_error = tryCatch(tolower(x), error=function(e) e)
                     # if not an error
                     if (!inherits(try_error, "error"))
                       y = tolower(x)
                     # result
                     return(y)
                   }
                   # use tryTolower with sapply 
                   sentence = sapply(sentence, tryTolower)
                   
                   # split sentence into words with str_split (stringr package)
                   word.list = str_split(sentence, "\\s+")
                   words = unlist(word.list)
                   
                   # compare words to the dictionaries of positive & negative terms
                   pos.matches = match(words, pos.words)
                   neg.matches = match(words, neg.words)
                   
                   # get the position of the matched term or NA
                   # we just want a TRUE/FALSE
                   pos.matches = !is.na(pos.matches)
                   neg.matches = !is.na(neg.matches)
                   
                   # final score
                   score = sum(pos.matches) - sum(neg.matches)
                   return(score)
                 }, pos.words, neg.words, .progress=.progress )
  
  # data frame with scores for each sentence
  scores.df = data.frame(text=sentences, score=scores)
  return(scores.df)
}

#  IMPORT POSITIVE AND NEGATIVE WORDS
pos.words = scan("positive_words.txt", what = 'character', comment.char = ";")
neg.words = scan("negative_words.txt", what = 'character', comment.char = ";")
moditwit <- read.csv("ModiTwits.csv",sep=",")
moditwit$text <- as.factor(moditwit$text)
modi.score <- score.sentiment(moditwit$text, pos.words, neg.words, .progress ='text')

write.csv(modi.score, "ModiScore.csv", row.names = F)

## PLOTTING THE OUTCOME
hist(modi.score$score, xlab = "Score of Tweets", main = "Frequency of Twits sentiment")



########################### MODULE 3 ################################
############### TEXT MINING AND WORDCLOUND DRAWING IN R #############

library(tm)
library(wordcloud)
#moditwit <- path  #assigning the directory where only moditwit file is saved

modiTwitPath <- "E:/R Programs/Twitter Sentiment Analysis/TwitData"

moditwit = Corpus(DirSource(modiTwitPath), readerControl = list(Language = "eng"))

## DOING NORMAL DATA FORMATTING

moditwit <- tm_map(moditwit, tolower)
moditwit <- tm_map(moditwit, stripWhitespace)
moditwit <- tm_map(moditwit, removePunctuation)
moditwit <- tm_map(moditwit, removeWords, c ("Modi"))

tdm <- TermDocumentMatrix(moditwit)
m1 <- as.matrix (tdm)
v1 <- sort(rowSums(m1), decreasing = TRUE)
d1 <- data.frame(word = names(v1), freq = v1)
wordcloud(d1$word, d1$freq, col=brewer.pal(8, "Set2"), min.freq = 1)



########################### THE END  ################################
############ TWITTER SENTIMENT ANALYSIS CODE BY DEEPESH #############