Further analysis on conistancy of sentiment has been done.

load("D:/Hohenheim/SEM 3/ADS/introads_ass2_team18/01_data/raw/gamereviews.RData")

# Only keeping necessary columns for analysis

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

rev.sentiment <- data.frame(id= reviews.clean$recommendationid,  reviews = reviews.clean$reviews)

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

cat("An emotion with highest weightage is disgust with", round(nrc.scores$weightage[10],2),"%\n")

# Checking the consistency of sentiment with positive and negative voted_up

merged <- merge(sent.label= rev.sentiment$sent.label, voted_up= gamereviews$voted_up, by.x = "id", 

                by.y = "recommendationid", all.x = TRUE)

comparison <- data.frame(id = merged$id, reviews = merged$reviews, sent.label=merged$sent.label, voted_up=merged$voted_up)

comparison <- comparison %>%

  mutate(voted_up_mapped = case_when(

    voted_up==TRUE ~ "Positive",

    voted_up==FALSE ~ "Negative"))

comparison$compare <- ifelse(comparison$sent.label == "Positive" & comparison$voted_up_mapped == "Positive", "positive",

                             ifelse(comparison$sent.label == "Negative" & comparison$voted_up_mapped == "Negative", "positive",

                                    "Neutral"))

# Counting the number of match 

match.count <- sum(comparison$compare=="Match")

mismatch.count <- sum(comparison$compare=="Mismatch")

total.count <- length(comparison$compare)

# Calculating ration of match and mismatch

match.ratio <- match.count / total.count

# Display the result

print(comparison_result)