Script for data analysis made to chcek sentiment of a review.

rm(list = ls())

# Installation of packages

install.packages("stringr")

# Loading of packages

library(stringr)

#loading necessary dataset

setwd("D:/Hohenheim/SEM 3/ADS/introads_ass2_team18/01_data")

reviews.clean <- read.csv("reviews.clean.csv")

# Only keeping necessary columns for analysis

rev.sentiment <- data.frame(reviews = reviews.clean$review)

# Manual sentiment analysis-----------------------------------------------------

# Loading of necessary words lists taken from Kaggel by Hu and Bing Liu

setwd("D:/Hohenheim/SEM 3/ADS/introads_ass2_team18/00_docs")

positive.words <- readLines("positive-words.txt")

negative.words <- readLines("negative-words.txt")

# Function to perform sentiment analysis on a single review; in a code, sent=sentiment

sent.analysis <- function(review) {

  # Tokenize the review into words

  words <- unlist(strsplit(review, "\\s+"))

  # Initialize sentiment score

  sent.score <- 0

  # Loop through each word in the review

  for (word in words) {

    # Check if the word is in the positive or negative word lists

    if (word %in% positive.words) {

      sent.score <- sent.score + 1

    } else if (word %in% negative.words) {

      sent.score <- sent.score - 1

    }

  }

  # Check for negation (not) before an opinion word

  if ("not" %in% words) {

    sent.score <- sent.score * -1

  }

  # Ascribe positive (1), negative (-1), or neutral (0) sentiment

  sentiment <- ifelse(sent.score > 0, 1, ifelse(sent.score < 0, -1, 0))

  return(sentiment)

}

# Apply sentiment analysis to all reviews in the dataframe

rev.sentiment$sentiment.score <- sapply(rev.sentiment$reviews, sent.analysis)

# Labeling the sentiment score with sentiment

rev.sentiment$sentiment.label <- ifelse(rev.sentiment$sentiment.score == 1, "Positive",

                                        ifelse(rev.sentiment$sentiment.score == -1, "Negative", "Neutral"))