DATA607 - Assignment Week 10
Sentiment Analysis
Donald Butler
10/31/2021
Introduction
We were asked to recreate the R code from chapter 2 of the book “Text Mining with R: A Tidy Approach” then perform similar analysis on another authors set of works.
Load Packages
Load the packages required to complete the sentiment analysis.
library(tidyverse)
library(tidytext)
library(janeaustenr)
library(wordcloud)
library(reshape2)Recreate Sentiment Analysis
Recreate the sentiment analysis from Chapter 2 in “Text Mining with R: A Tidy Approach.” (Silge and Robinson 2017)
Sentiment Lexicons
A sentiment lexicon is a set of words and a sentiment score that is attributed to each word.
AFINN-111
The AFINN sentiment lexicon scores words on a scale of -5 (negative sentiment) to 5 (positive sentiment). (Nielsen 2011)
get_sentiments("afinn")## # A tibble: 2,477 x 2
## word value
## <chr> <dbl>
## 1 abandon -2
## 2 abandoned -2
## 3 abandons -2
## 4 abducted -2
## 5 abduction -2
## 6 abductions -2
## 7 abhor -3
## 8 abhorred -3
## 9 abhorrent -3
## 10 abhors -3
## # ... with 2,467 more rowsBing
The Bing sentiment lexicon assigns words as either positive or negative. (Hu and Liu 2004)
get_sentiments("bing")## # A tibble: 6,786 x 2
## word sentiment
## <chr> <chr>
## 1 2-faces negative
## 2 abnormal negative
## 3 abolish negative
## 4 abominable negative
## 5 abominably negative
## 6 abominate negative
## 7 abomination negative
## 8 abort negative
## 9 aborted negative
## 10 aborts negative
## # ... with 6,776 more rowsNRC Word-Emotion Association Lexicon
The nrc sentiment lexicon classifies words as positive or negative, but also by emotions such as fear or anger. (Mohammad and Turney 2013)
get_sentiments("nrc")## # A tibble: 13,875 x 2
## word sentiment
## <chr> <chr>
## 1 abacus trust
## 2 abandon fear
## 3 abandon negative
## 4 abandon sadness
## 5 abandoned anger
## 6 abandoned fear
## 7 abandoned negative
## 8 abandoned sadness
## 9 abandonment anger
## 10 abandonment fear
## # ... with 13,865 more rowsSentiment in Jane Austen’s Books
Using the Bing sentiment lexicon to compare the sentiment of all of Jane Austen’s books.
lines_per_index <- 80
tidy_books <- austen_books() %>%
group_by(book) %>%
mutate(
linenumber = row_number(),
chapter = cumsum(str_detect(text,regex("^chapter [\\divxlc]", ignore_case = TRUE)))
) %>%
ungroup() %>%
unnest_tokens(word,text)
jane_austen_sentiment <- tidy_books %>%
inner_join(get_sentiments('bing'), by = 'word') %>%
count(book, index = linenumber %/% lines_per_index, sentiment) %>%
pivot_wider(names_from = sentiment, values_from = n, values_fill = 0) %>%
mutate(sentiment = positive - negative)
jane_austen_sentiment %>%
ggplot(aes(index, sentiment, fill = book)) +
geom_col(show.legend = FALSE) +
facet_wrap(~book, ncol = 2, scales = "free_x")
Comparing Sentiment Lexicons
In Jane Austen’s book “Pride and Prejudice,” we will compare the three sentiment lexicons.
pride_prejudice <- tidy_books %>%
filter(book == "Pride & Prejudice")
afinn <- pride_prejudice %>%
inner_join(get_sentiments('afinn'), by = 'word') %>%
group_by(index = linenumber %/% lines_per_index) %>%
summarize(sentiment = sum(value)) %>%
mutate(method = "AFINN")
bing_and_nrc <- bind_rows(
pride_prejudice %>%
inner_join(get_sentiments('bing'), by = 'word') %>%
mutate(method = "Bing et al."),
pride_prejudice %>%
inner_join(get_sentiments('nrc') %>%
filter(sentiment %in% c('positive','negative')), by = 'word') %>%
mutate(method = 'NRC')
) %>%
count(method, index = linenumber %/% lines_per_index, sentiment) %>%
pivot_wider(names_from = sentiment, values_from = n, values_fill = 0) %>%
mutate(sentiment = positive - negative)
bind_rows(afinn, bing_and_nrc) %>%
ggplot(aes(index, sentiment, fill = method)) +
geom_col(show.legend = FALSE) +
facet_wrap(~method, ncol = 1, scales = 'free_y')
Common Positive and Negative Words
Determine the most common positive and negative words in Jane Austen’s books.
bing_word_counts <- tidy_books %>%
inner_join(get_sentiments('bing'), by = 'word') %>%
count(word, sentiment, sort = TRUE) %>%
ungroup()
bing_word_counts %>%
group_by(sentiment) %>%
slice_max(n, n = 10) %>%
ungroup() %>%
mutate(word = reorder(word, n)) %>%
ggplot(aes(n, word, fill = sentiment)) +
geom_col(show.legend = FALSE) +
facet_wrap(~sentiment, scales = 'free_y') +
labs(x = 'Contribution to sentiment', y = NULL)
Custom Stop Words
The word ‘miss’ is contributing significantly towards negative sentiment, but it’s use within Jane Austen’s books isn’t used in that context. Create a custom stop word list to adjust for anomalies in the sentiment lexicon used.
custom_stop_words <- bind_rows(tibble(word = c('miss'), lexicon = c('custom')), stop_words)
custom_stop_words## # A tibble: 1,150 x 2
## word lexicon
## <chr> <chr>
## 1 miss custom
## 2 a SMART
## 3 a's SMART
## 4 able SMART
## 5 about SMART
## 6 above SMART
## 7 according SMART
## 8 accordingly SMART
## 9 across SMART
## 10 actually SMART
## # ... with 1,140 more rowsWordclouds
Construct a wordcloud of the most common words in Jane Austen’s books.
tidy_books %>%
anti_join(stop_words, by = 'word') %>%
count(word) %>%
with(wordcloud(word, n, max.words = 100))
Wordcoulds by Sentiment
Construct a wordcloud of the positive and negative words in Jane Austen’s books.
tidy_books %>%
inner_join(get_sentiments('bing'), by = 'word') %>%
count(word, sentiment, sort = TRUE) %>%
acast(word ~ sentiment, value.var = 'n', fill = 0) %>%
comparison.cloud(colors = c('gray20', 'gray80'), max.words = 100)
Sentiment based on other units
Instead of considering sentiment on a word for word basis, we can look at it based on other units of text such as chapters. Determine the most negative chapter in each book.
austen_chapters <- austen_books() %>%
group_by(book) %>%
unnest_tokens(chapter, text, token = 'regex', pattern = "Chapter|CHAPTER|chapter [\\divxlc]") %>%
ungroup()
austen_chapters %>%
group_by(book) %>%
summarize(chapters = n())## # A tibble: 6 x 2
## book chapters
## <fct> <int>
## 1 Sense & Sensibility 51
## 2 Pride & Prejudice 62
## 3 Mansfield Park 49
## 4 Emma 56
## 5 Northanger Abbey 32
## 6 Persuasion 25bing_negative <- get_sentiments('bing') %>%
filter(sentiment == 'negative')
wordcounts <- tidy_books %>%
group_by(book, chapter) %>%
summarize(words = n())## `summarise()` has grouped output by 'book'. You can override using the `.groups` argument.tidy_books %>%
semi_join(bing_negative, by = 'word') %>%
group_by(book, chapter) %>%
summarize(negativewords = n()) %>%
left_join(wordcounts, by = c('book','chapter')) %>%
mutate(ratio = negativewords / words) %>%
filter(chapter != 0) %>%
slice_max(ratio, n = 1) %>%
ungroup()## `summarise()` has grouped output by 'book'. You can override using the `.groups` argument.## # A tibble: 6 x 5
## book chapter negativewords words ratio
## <fct> <int> <int> <int> <dbl>
## 1 Sense & Sensibility 43 161 3405 0.0473
## 2 Pride & Prejudice 34 111 2104 0.0528
## 3 Mansfield Park 46 173 3685 0.0469
## 4 Emma 15 151 3340 0.0452
## 5 Northanger Abbey 21 149 2982 0.0500
## 6 Persuasion 4 62 1807 0.0343Sir Arthur Conan Doyle
Perform a similar analysis based on the works of author Sir Arthur Conan Doyle.
Project Gutenberg
Project Gutenberg is a library of public domain works. The gutenbergr package allows access to these works. (Robinson 2021) Retrieve a list of the Sherlock Holmes books that are available.
library(gutenbergr)
doyle_books <- gutenberg_works(author == 'Doyle, Arthur Conan') %>%
filter(gutenberg_bookshelf == 'Detective Fiction')
doyle_books## # A tibble: 13 x 8
## gutenberg_id title author gutenberg_autho~ language gutenberg_books~ rights
## <int> <chr> <chr> <int> <chr> <chr> <chr>
## 1 108 The Re~ Doyle~ 69 en Detective Ficti~ Publi~
## 2 244 A Stud~ Doyle~ 69 en Detective Ficti~ Publi~
## 3 834 The Me~ Doyle~ 69 en Detective Ficti~ Publi~
## 4 2097 The Si~ Doyle~ 69 en Detective Ficti~ Publi~
## 5 2343 The Ad~ Doyle~ 69 en Detective Ficti~ Publi~
## 6 2344 The Ad~ Doyle~ 69 en Detective Ficti~ Publi~
## 7 2345 The Ad~ Doyle~ 69 en Detective Ficti~ Publi~
## 8 2346 The Ad~ Doyle~ 69 en Detective Ficti~ Publi~
## 9 2347 The Ad~ Doyle~ 69 en Detective Ficti~ Publi~
## 10 2348 The Di~ Doyle~ 69 en Detective Ficti~ Publi~
## 11 2349 The Ad~ Doyle~ 69 en Detective Ficti~ Publi~
## 12 2350 His La~ Doyle~ 69 en Detective Ficti~ Publi~
## 13 3289 The Va~ Doyle~ 69 en Detective Ficti~ Publi~
## # ... with 1 more variable: has_text <lgl>With the list of books, download the works and tidy.
tidy_doyle <- doyle_books %>%
gutenberg_download(meta_fields = 'title') %>%
group_by(gutenberg_id) %>%
mutate(linenumber = row_number()) %>%
ungroup() %>%
unnest_tokens(word, text)## Determining mirror for Project Gutenberg from http://www.gutenberg.org/robot/harvest## Using mirror http://aleph.gutenberg.orgUnfortunately, in this collection of works, chapter pages are not present, so we are unable to consider a units by chapter.
Sentiment in Sir Arthur Conan Doyal’s Books
Using the Bing sentiment lexicon to compare the sentiment of all of the Sherlock Holmes books.
lines_per_index <- 80
doyle_sentiment <- tidy_doyle %>%
inner_join(get_sentiments('bing'), by = 'word') %>%
count(title, index = linenumber %/% lines_per_index, sentiment) %>%
pivot_wider(names_from = sentiment, values_from = n, values_fill = 0) %>%
mutate(sentiment = positive - negative)
doyle_sentiment %>%
ggplot(aes(index, sentiment, fill = title)) +
geom_col(show.legend = FALSE) +
facet_wrap(~title, ncol = 3, scales = "free_x")
Comparing Sentiment Lexicons
In “A Study in Scarlet,” we will compare the three sentiment lexicons.
study_scarlet <- tidy_doyle %>%
filter(title == "A Study in Scarlet")
scarlet_afinn <- study_scarlet %>%
inner_join(get_sentiments('afinn'), by = 'word') %>%
group_by(index = linenumber %/% lines_per_index) %>%
summarize(sentiment = sum(value)) %>%
mutate(method = "AFINN")
scarlet_bing_and_nrc <- bind_rows(
study_scarlet %>%
inner_join(get_sentiments('bing'), by = 'word') %>%
mutate(method = "Bing et al."),
study_scarlet %>%
inner_join(get_sentiments('nrc') %>%
filter(sentiment %in% c('positive','negative')), by = 'word') %>%
mutate(method = 'NRC')
) %>%
count(method, index = linenumber %/% lines_per_index, sentiment) %>%
pivot_wider(names_from = sentiment, values_from = n, values_fill = 0) %>%
mutate(sentiment = positive - negative)
bind_rows(scarlet_afinn, scarlet_bing_and_nrc) %>%
ggplot(aes(index, sentiment, fill = method)) +
geom_col(show.legend = FALSE) +
facet_wrap(~method, ncol = 1, scales = 'free_y')
Syuzhet
I tried playing with the Syuzhet package to use another sentiment lexicon. (Jockers 2015) The package computes a sentiment score for each line of the book text.
library(syuzhet)
study_in_scarlet <- doyle_books %>%
filter(title == 'A Study in Scarlet') %>%
gutenberg_download(meta_fields = 'title') %>%
filter(grepl('[A-Za-z]',text))
syuzhet_sentiment <- get_sentiment(study_in_scarlet$text, method = 'syuzhet')
summary(syuzhet_sentiment)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -3.60000 -0.40000 0.00000 0.01536 0.50000 2.60000Common Positive and Negative Words
Determine the most common positive and negative words in the Sherlock Holmes books.
doyle_bing_word_counts <- tidy_doyle %>%
inner_join(get_sentiments('bing'), by = 'word') %>%
count(word, sentiment, sort = TRUE) %>%
ungroup()
doyle_bing_word_counts %>%
group_by(sentiment) %>%
slice_max(n, n = 10) %>%
ungroup() %>%
mutate(word = reorder(word, n)) %>%
ggplot(aes(n, word, fill = sentiment)) +
geom_col(show.legend = FALSE) +
facet_wrap(~sentiment, scales = 'free_y') +
labs(x = 'Contribution to sentiment', y = NULL)
Custom Stop Words
Again we see the word ‘miss’ on the common negative list, although not as impacting with Jane Austen’s work, but based on the time period I would expect a fair number of instances are referring to a women and not the negative context.
Wordclouds
Construct a wordcloud of the most common words in the Sherlock Holmes books.
tidy_doyle %>%
anti_join(stop_words, by = 'word') %>%
count(word) %>%
with(wordcloud(word, n, max.words = 100))
Wordcoulds by Sentiment
Construct a wordcloud of the positive and negative words in Jane Austen’s books.
tidy_doyle %>%
inner_join(get_sentiments('bing'), by = 'word') %>%
count(word, sentiment, sort = TRUE) %>%
acast(word ~ sentiment, value.var = 'n', fill = 0) %>%
comparison.cloud(colors = c('gray20', 'gray80'), max.words = 100)
Sentiment based on other units
Unfortunately the chapter pages that were in Jane Austen’s books, do not appear in the files for Sir Arthur Conan Doyle, so a similar analysis cannot be performed.