Analyzing customer review of a product by text mining to find most frequent words to know the trend.
library("tm")
## Warning: package 'tm' was built under R version 4.1.2
## Loading required package: NLP
## Warning: package 'NLP' was built under R version 4.1.1
library("SnowballC")
## Warning: package 'SnowballC' was built under R version 4.1.1
library("wordcloud")
## Warning: package 'wordcloud' was built under R version 4.1.2
## Loading required package: RColorBrewer
## Warning: package 'RColorBrewer' was built under R version 4.1.1
library("RColorBrewer")
text <- readLines("C:/Users/Punalur/Documents/abc.txt")
head(text, 3)
## [1] "This case serves it's function well - absorbs falls, it's grippy soft (but hard) and holds the phone in place well. This is the second one of these I've purchased, so I must be satisfied. But it also means the last one eventually broke. I can say I used the last one over a year and dropped my phone several times, never taking any damage. The front edge is raised just enough to protect the screen. Eventually, the bottom corner of the case cracked and, although it never completely separated, I figured I'd change the case before pushing my luck further."
## [2] "I would have bought the exact same red one, but color options are dwindling. I settled for mint and it was a little lighter than I expected. I'd call the color \"robin egg\", not mint. M'eh"
## [3] "Great little phone case, not too bulky and the colour is great! The only things I'm not thrilled about: it's a very slippery phone case and the case doesn't fully cover the bottom of the phone, so if you drop it you could still damage your phone. Otherwise, super cute!"
document <- Corpus(VectorSource(text))
# cleaning the text
document <- tm_map(document, content_transformer(tolower)) # change to lowercase
## Warning in tm_map.SimpleCorpus(document, content_transformer(tolower)):
## transformation drops documents
document <- tm_map(document, removeNumbers) # remove numbers
## Warning in tm_map.SimpleCorpus(document, removeNumbers): transformation drops
## documents
document <- tm_map(document, removeWords, stopwords("english")) # remove stopwords in English
## Warning in tm_map.SimpleCorpus(document, removeWords, stopwords("english")):
## transformation drops documents
document <- tm_map(document, removeWords, c("also")) # specify own stopwords to be removed
## Warning in tm_map.SimpleCorpus(document, removeWords, c("also")): transformation
## drops documents
# Build table withthe words and their frequency
dtm <- TermDocumentMatrix(document)
m <- as.matrix(dtm) # Error: cannot allocate vector of size 39.1 Gb
v <- sort(rowSums(m), decreasing = TRUE)
D <- data.frame(word = names(v), frequency = v)
head(D, 10)
## word frequency
## case case 15
## phone phone 9
## colour colour 7
## nice nice 7
## got got 5
## enough enough 4
## just just 4
## protect protect 4
## well well 4
## little little 4
# Create word cloud
set.seed(1234)
wordcloud(words = D$word, freq = D$frequency, min.freq = 1,
max.words = 20, random.order = FALSE, rot.per = 0.05,
colors = brewer.pal(8, "Dark2"))
# Examin frequent terms and their association
findAssocs(dtm, terms = "phone", corlimit = 0.2)
## $phone
## case bottom well. super protect
## 0.73 0.67 0.67 0.67 0.46
## absorbs although broke. change completely
## 0.45 0.45 0.45 0.45 0.45
## corner damage. edge eventually eventually,
## 0.45 0.45 0.45 0.45 0.45
## falls, figured front function grippy
## 0.45 0.45 0.45 0.45 0.45
## hard) holds last luck means
## 0.45 0.45 0.45 0.45 0.45
## must never one place purchased,
## 0.45 0.45 0.45 0.45 0.45
## pushing raised satisfied. say screen.
## 0.45 0.45 0.45 0.45 0.45
## second separated, serves several soft
## 0.45 0.45 0.45 0.45 0.45
## taking used year bulky cover
## 0.45 0.45 0.45 0.45 0.45
## cute! damage drop fully great
## 0.45 0.45 0.45 0.45 0.45
## great! phone. still things thrilled
## 0.45 0.45 0.45 0.45 0.45
## (clove absolutely amazon. another better
## 0.45 0.45 0.45 0.45 0.45
## fantastic! feel justice. person, photos
## 0.45 0.45 0.45 0.45 0.45
## pictured purple) purple. quality quality!
## 0.45 0.45 0.45 0.45 0.45
## sturdier sturdy/strong way yet dropped
## 0.45 0.45 0.45 0.45 0.39
## enough times, phone, slippery exactly
## 0.39 0.39 0.39 0.39 0.39
## feels cracked silicone light colour
## 0.39 0.27 0.27 0.27 0.22
# find frequent terms
findFreqTerms(dtm, lowfreq = 2)
## [1] "bottom" "can" "case" "cracked" "dropped"
## [6] "enough" "just" "last" "never" "one"
## [11] "phone" "protect" "times," "well" "well."
## [16] "color" "little" "mint" "case," "colour"
## [21] "otherwise," "phone," "slippery" "super" "case."
## [26] "got" "nice" "really" "silicone" "exactly"
## [31] "feels" "like" "love" "protection" "orange"
## [36] "red." "good" "slip" "light" "pretty"
## [41] "screen"
# Plot most frequent words
barplot(D[1:10,]$frequency, las = 2, names.arg = D[1:10,]$word,
col = "lightblue", main = "Most frequent words",
ylab = "Word frequencies")
