What factors are most influential in enhancing customer satisfaction for tourists staying in hotels in Paris, France?
Heeseo Yun, Youbin Hwang
December 11, 2024
load library
# Add any libraries and general settings up here.
# I suggest you start with these two libraries, since you'll probably use them:
library(tidyverse)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.1
## ✔ ggplot2 3.5.1 ✔ tibble 3.2.1
## ✔ lubridate 1.9.3 ✔ tidyr 1.3.1
## ✔ purrr 1.0.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errorslibrary(here)## here() starts at /Users/heeseoyun/Downloads/Final_example (1)library(dplyr)
library(readr)
library(ggplot2)
library(wordcloud)## Loading required package: RColorBrewerlibrary(tidytext)
library(tidyr)
library(stringr)
library(RColorBrewer)Executive summary
What is (are) your main question(s)? What is your story? What does the final graphic show?
This research seeks to answer the following question: “What factors are most influential in enhancing customer satisfaction for tourists staying in hotels in Paris, France?”
The study focuses on uncovering these factors using 515,000 customer reviews from Booking.com. By examining textual (positive/negative reviews) and numerical (ratings and word counts) data, it identifies the key drivers of satisfaction and dissatisfaction. The insights aim to support the strategic decision-making of hoteliers by highlighting areas that need improvement and reinforcing successful aspects of their services.
Techniques like keyword extraction, sentiment analysis, and visualization tools are employed. Word clouds provide a high-level view of frequently mentioned keywords, while bar graphs and scatter plots analyze relationships between satisfaction factors and average scores.
By addressing these questions, this research contributes to the sustainable growth of the hospitality industry in Paris, especially post-pandemic. Hotels can leverage these findings to enhance customer satisfaction, strengthen market competitiveness, and meet evolving expectations for service quality.
Data background
Explain where the data came from, what agency or company made it, how it is structured, what it shows, etc.
The dataset used in this study, “515K Hotel Reviews Data in Europe,” from kaggle employs the 515K Hotel Reviews Data in Europe, sourced from Booking.com. The dataset contains over 515,000 reviews and ratings for 1,493 luxury hotels, offering both numerical and textual insights into customer satisfaction.
Selected Variables and Their Relevance: These variables were selected to balance quantitative and qualitative insights. Numerical scores help quantify satisfaction levels, while positive and negative reviews uncover the specific reasons behind customer opinions. By focusing on these elements, the analysis can comprehensively address the research question.
Hotel_Name: This variable is essential for identifying specific hotels within the dataset. By retaining this information, we can analyze satisfaction factors at the individual hotel level, allowing for meaningful comparisons across different hotels. This enables us to pinpoint which hotels excel or underperform based on customer feedback and satisfaction metrics.
Average_Score: The Average_Score represents the overall satisfaction score aggregated across all reviews for a given hotel. It serves as a critical indicator of hotel performance and provides a benchmark for comparing satisfaction levels across hotels. This variable helps us identify trends in customer satisfaction and how other factors, such as keyword ratios, correlate with overall ratings.
Positive_Review: The Positive_Review variable contains the text of favorable comments left by customers. By analyzing this variable, we can identify the aspects of a hotel that customers appreciate most, such as cleanliness, service quality, or location. These insights highlight the strengths that hotels should emphasize to maintain and enhance guest satisfaction.
Negative_Review: The Negative_Review variable contains the text of unfavorable comments. This variable is critical for identifying common complaints or issues that lead to customer dissatisfaction. Analyzing this feedback provides actionable insights into areas that require improvement, such as breakfast quality, pricing, or amenities.
Data Cleaning
Describe and show how you cleaned and reshaped the data
Load Data
Load 515K Hotel Reviews Data in Europe. The data is imported for analysis and verified by displaying the first few rows.
# Load the data using relative path
hotel_reviews <- read.csv("Hotel_Reviews.csv", stringsAsFactors = FALSE)
# View the first few rows to verify
head(hotel_reviews)## Hotel_Address
## 1 s Gravesandestraat 55 Oost 1092 AA Amsterdam Netherlands
## 2 s Gravesandestraat 55 Oost 1092 AA Amsterdam Netherlands
## 3 s Gravesandestraat 55 Oost 1092 AA Amsterdam Netherlands
## 4 s Gravesandestraat 55 Oost 1092 AA Amsterdam Netherlands
## 5 s Gravesandestraat 55 Oost 1092 AA Amsterdam Netherlands
## 6 s Gravesandestraat 55 Oost 1092 AA Amsterdam Netherlands
## Additional_Number_of_Scoring Review_Date Average_Score Hotel_Name
## 1 194 8/3/2017 7.7 Hotel Arena
## 2 194 8/3/2017 7.7 Hotel Arena
## 3 194 7/31/2017 7.7 Hotel Arena
## 4 194 7/31/2017 7.7 Hotel Arena
## 5 194 7/24/2017 7.7 Hotel Arena
## 6 194 7/24/2017 7.7 Hotel Arena
## Reviewer_Nationality
## 1 Russia
## 2 Ireland
## 3 Australia
## 4 United Kingdom
## 5 New Zealand
## 6 Poland
## Negative_Review
## 1 I am so angry that i made this post available via all possible sites i use when planing my trips so no one will make the mistake of booking this place I made my booking via booking com We stayed for 6 nights in this hotel from 11 to 17 July Upon arrival we were placed in a small room on the 2nd floor of the hotel It turned out that this was not the room we booked I had specially reserved the 2 level duplex room so that we would have a big windows and high ceilings The room itself was ok if you don t mind the broken window that can not be closed hello rain and a mini fridge that contained some sort of a bio weapon at least i guessed so by the smell of it I intimately asked to change the room and after explaining 2 times that i booked a duplex btw it costs the same as a simple double but got way more volume due to the high ceiling was offered a room but only the next day SO i had to check out the next day before 11 o clock in order to get the room i waned to Not the best way to begin your holiday So we had to wait till 13 00 in order to check in my new room what a wonderful waist of my time The room 023 i got was just as i wanted to peaceful internal garden view big window We were tired from waiting the room so we placed our belongings and rushed to the city In the evening it turned out that there was a constant noise in the room i guess it was made by vibrating vent tubes or something it was constant and annoying as hell AND it did not stop even at 2 am making it hard to fall asleep for me and my wife I have an audio recording that i can not attach here but if you want i can send it via e mail The next day the technician came but was not able to determine the cause of the disturbing sound so i was offered to change the room once again the hotel was fully booked and they had only 1 room left the one that was smaller but seems newer
## 2 No Negative
## 3 Rooms are nice but for elderly a bit difficult as most rooms are two story with narrow steps So ask for single level Inside the rooms are very very basic just tea coffee and boiler and no bar empty fridge
## 4 My room was dirty and I was afraid to walk barefoot on the floor which looked as if it was not cleaned in weeks White furniture which looked nice in pictures was dirty too and the door looked like it was attacked by an angry dog My shower drain was clogged and the staff did not respond to my request to clean it On a day with heavy rainfall a pretty common occurrence in Amsterdam the roof in my room was leaking luckily not on the bed you could also see signs of earlier water damage I also saw insects running on the floor Overall the second floor of the property looked dirty and badly kept On top of all of this a repairman who came to fix something in a room next door at midnight was very noisy as were many of the guests I understand the challenges of running a hotel in an old building but this negligence is inconsistent with prices demanded by the hotel On the last night after I complained about water damage the night shift manager offered to move me to a different room but that offer came pretty late around midnight when I was already in bed and ready to sleep
## 5 You When I booked with your company on line you showed me pictures of a room I thought I was getting and paying for and then when we arrived that s room was booked and the staff told me we could only book the villa suite theough them directly Which was completely false advertising After being there we realised that you have grouped lots of rooms on the photos together leaving me the consumer confused and extreamly disgruntled especially as its my my wife s 40th birthday present Please make your website more clear through pricing and photos as again I didn t really know what I was paying for and how much it had wnded up being Your photos told me I was getting something I wasn t Not happy and won t be using you again
## 6 Backyard of the hotel is total mess shouldn t happen in hotel with 4 stars
## Review_Total_Negative_Word_Counts Total_Number_of_Reviews
## 1 397 1403
## 2 0 1403
## 3 42 1403
## 4 210 1403
## 5 140 1403
## 6 17 1403
## Positive_Review
## 1 Only the park outside of the hotel was beautiful
## 2 No real complaints the hotel was great great location surroundings rooms amenities and service Two recommendations however firstly the staff upon check in are very confusing regarding deposit payments and the staff offer you upon checkout to refund your original payment and you can make a new one Bit confusing Secondly the on site restaurant is a bit lacking very well thought out and excellent quality food for anyone of a vegetarian or vegan background but even a wrap or toasted sandwich option would be great Aside from those minor minor things fantastic spot and will be back when i return to Amsterdam
## 3 Location was good and staff were ok It is cute hotel the breakfast range is nice Will go back
## 4 Great location in nice surroundings the bar and restaurant are nice and have a lovely outdoor area The building also has quite some character
## 5 Amazing location and building Romantic setting
## 6 Good restaurant with modern design great chill out place Great park nearby the hotel and awesome main stairs
## Review_Total_Positive_Word_Counts Total_Number_of_Reviews_Reviewer_Has_Given
## 1 11 7
## 2 105 7
## 3 21 9
## 4 26 1
## 5 8 3
## 6 20 1
## Reviewer_Score
## 1 2.9
## 2 7.5
## 3 7.1
## 4 3.8
## 5 6.7
## 6 6.7
## Tags
## 1 [' Leisure trip ', ' Couple ', ' Duplex Double Room ', ' Stayed 6 nights ']
## 2 [' Leisure trip ', ' Couple ', ' Duplex Double Room ', ' Stayed 4 nights ']
## 3 [' Leisure trip ', ' Family with young children ', ' Duplex Double Room ', ' Stayed 3 nights ', ' Submitted from a mobile device ']
## 4 [' Leisure trip ', ' Solo traveler ', ' Duplex Double Room ', ' Stayed 3 nights ']
## 5 [' Leisure trip ', ' Couple ', ' Suite ', ' Stayed 2 nights ', ' Submitted from a mobile device ']
## 6 [' Leisure trip ', ' Group ', ' Duplex Double Room ', ' Stayed 1 night ']
## days_since_review lat lng
## 1 0 days 52.36058 4.915968
## 2 0 days 52.36058 4.915968
## 3 3 days 52.36058 4.915968
## 4 3 days 52.36058 4.915968
## 5 10 days 52.36058 4.915968
## 6 10 days 52.36058 4.915968Filltering Data 1: Paris Hotel
To focus the analysis on Paris, the dataset is filtered to include only hotels located in this city. Rows where the Hotel_Address column contains the word “Paris” are extracted. This reduces the dataset size, optimizing computation and ensuring relevance to the research question.
# Filter the data for rows where the Hotel_Address contains "Paris"
paris_hotels <- hotel_reviews %>%
filter(grepl("Paris", Hotel_Address, ignore.case = TRUE))
# View the first few rows of the filtered data
head(paris_hotels)## Hotel_Address Additional_Number_of_Scoring
## 1 1 3 Rue d Argentine 16th arr 75116 Paris France 26
## 2 1 3 Rue d Argentine 16th arr 75116 Paris France 26
## 3 1 3 Rue d Argentine 16th arr 75116 Paris France 26
## 4 1 3 Rue d Argentine 16th arr 75116 Paris France 26
## 5 1 3 Rue d Argentine 16th arr 75116 Paris France 26
## 6 1 3 Rue d Argentine 16th arr 75116 Paris France 26
## Review_Date Average_Score Hotel_Name Reviewer_Nationality
## 1 6/29/2017 8.4 Monhotel Lounge SPA Brazil
## 2 4/25/2017 8.4 Monhotel Lounge SPA Luxembourg
## 3 4/18/2017 8.4 Monhotel Lounge SPA United Kingdom
## 4 11/11/2016 8.4 Monhotel Lounge SPA Belgium
## 5 8/28/2016 8.4 Monhotel Lounge SPA Qatar
## 6 8/28/2016 8.4 Monhotel Lounge SPA Saudi Arabia
## Negative_Review
## 1 No Negative
## 2 Not only did the staff on arrival ask to copy the details of my credit card this is OK but they also noted down the CVV number from the back of the credit card This means that the hotel has all the information to use my credit card for internet purchases Since this is unacceptable I asked the receptionist to delete the CVV but she refused This policy should be changed Otherwise good hotel
## 3 N A
## 4 The sauna is a wonderfull addition it s a shame I had to slalom in the cleaning staff halway almost filled with cleaning products Also the proximity to the cleaning staff s room added all the noise to what was supposed to be a quiet moment maybe it would be a good idea to relocate their general quarter
## 5 They have maintenance and I couldn t enjoy the balcony view
## 6 Reception not impressive building was under construction Room passage was way to small Room has also no space
## Review_Total_Negative_Word_Counts Total_Number_of_Reviews
## 1 0 171
## 2 76 171
## 3 3 171
## 4 60 171
## 5 13 171
## 6 19 171
## Positive_Review
## 1 Nice hotel Room was beautiful and bed very comfortable Did not expect big rooms in Paris so size was really ok with 2 windows Bathroom modern and new with a GREAT shower Staff was wonderfull since reservation They sent me an email prior to my arrival asking if they could help me in anyway and also asking if I had any special needs such as extra beds I had a last minute health problem in the family and had to arrive one day earlier and stay for 1 night less Audrey was wonderful and managed to receive me before and dealt herself with Booking to change my reservation I sent a note to her when I was entering my flight from Cannes to Paris and when I left the airplane she had already sorted everything out They even charged me a lower rate as on the day I arrived the room costed a little less In this same day Saturday June 25 I went out with friends and arrived really hungry at the hotel at 1 30AM and the guy from reception can t believe I forgot his name Maybe Hadesmi managed to open the restaurant and get to my room a nice aspargus risoto bread yougurt and a warm smile even thought riom service finishes at midnight Really had a very nice experience at Monhotel and recomend it
## 2 No Positive
## 3 Fantastic weekend with my partner We Would definitely stay here again Location was excellent and couldn t have asked for a better area to stay in Was so close to the metro and all the shops bars and restaurants Room was lovely and cosey and very modern with a classy edge to the facility
## 4 Good location really good breakfast maybe the best I ever had in Paris in a confortable and stylish setting Everything you need is provided on site you can literally arrive with your hands in your pockets Helpfull and attentive staff
## 5 I asked to be downgraded because I was paying for a balcony room while I couldn t enjoy the balcony Instead they upgraded me with the biggest room in the hotel for the same rate
## 6 The room has everything inside as a full package
## Review_Total_Positive_Word_Counts Total_Number_of_Reviews_Reviewer_Has_Given
## 1 231 1
## 2 0 1
## 3 56 4
## 4 42 14
## 5 37 4
## 6 10 5
## Reviewer_Score
## 1 9.2
## 2 8.8
## 3 7.9
## 4 9.6
## 5 8.8
## 6 7.5
## Tags
## 1 [' Leisure trip ', ' Group ', ' Comfort Double Room ', ' Stayed 1 night ', ' Submitted from a mobile device ']
## 2 [' Business trip ', ' Solo traveler ', ' Superior Double Room ', ' Stayed 4 nights ']
## 3 [' Leisure trip ', ' Couple ', ' Premium Double Room ', ' Stayed 2 nights ', ' Submitted from a mobile device ']
## 4 [' Business trip ', ' Solo traveler ', ' Superior Double Room ', ' Stayed 4 nights ', ' Submitted from a mobile device ']
## 5 [' Business trip ', ' Solo traveler ', ' Deluxe Double Room with Balcony ', ' Stayed 5 nights ', ' Submitted from a mobile device ']
## 6 [' Leisure trip ', ' Solo traveler ', ' Premium Double Room ', ' Stayed 1 night ', ' Submitted from a mobile device ']
## days_since_review lat lng
## 1 35 days 48.87435 2.289733
## 2 100 day 48.87435 2.289733
## 3 107 day 48.87435 2.289733
## 4 265 day 48.87435 2.289733
## 5 340 day 48.87435 2.289733
## 6 340 day 48.87435 2.289733# Optional: Check the number of rows in the filtered data
nrow(paris_hotels)## [1] 59928Filltering Data 2: select and reorder data
From the filtered dataset, only the variables necessary for the study are selected. This includes variables representing average scores, hotel names, and review texts. Redundant variables are removed to streamline the dataset, and columns are reordered for convenience.
# Create a new data frame with only the required variables
selected_data <- paris_hotels %>%
select(Average_Score, Hotel_Name, Positive_Review, Negative_Review)
# View the first few rows of the new data frame
head(selected_data)## Average_Score Hotel_Name
## 1 8.4 Monhotel Lounge SPA
## 2 8.4 Monhotel Lounge SPA
## 3 8.4 Monhotel Lounge SPA
## 4 8.4 Monhotel Lounge SPA
## 5 8.4 Monhotel Lounge SPA
## 6 8.4 Monhotel Lounge SPA
## Positive_Review
## 1 Nice hotel Room was beautiful and bed very comfortable Did not expect big rooms in Paris so size was really ok with 2 windows Bathroom modern and new with a GREAT shower Staff was wonderfull since reservation They sent me an email prior to my arrival asking if they could help me in anyway and also asking if I had any special needs such as extra beds I had a last minute health problem in the family and had to arrive one day earlier and stay for 1 night less Audrey was wonderful and managed to receive me before and dealt herself with Booking to change my reservation I sent a note to her when I was entering my flight from Cannes to Paris and when I left the airplane she had already sorted everything out They even charged me a lower rate as on the day I arrived the room costed a little less In this same day Saturday June 25 I went out with friends and arrived really hungry at the hotel at 1 30AM and the guy from reception can t believe I forgot his name Maybe Hadesmi managed to open the restaurant and get to my room a nice aspargus risoto bread yougurt and a warm smile even thought riom service finishes at midnight Really had a very nice experience at Monhotel and recomend it
## 2 No Positive
## 3 Fantastic weekend with my partner We Would definitely stay here again Location was excellent and couldn t have asked for a better area to stay in Was so close to the metro and all the shops bars and restaurants Room was lovely and cosey and very modern with a classy edge to the facility
## 4 Good location really good breakfast maybe the best I ever had in Paris in a confortable and stylish setting Everything you need is provided on site you can literally arrive with your hands in your pockets Helpfull and attentive staff
## 5 I asked to be downgraded because I was paying for a balcony room while I couldn t enjoy the balcony Instead they upgraded me with the biggest room in the hotel for the same rate
## 6 The room has everything inside as a full package
## Negative_Review
## 1 No Negative
## 2 Not only did the staff on arrival ask to copy the details of my credit card this is OK but they also noted down the CVV number from the back of the credit card This means that the hotel has all the information to use my credit card for internet purchases Since this is unacceptable I asked the receptionist to delete the CVV but she refused This policy should be changed Otherwise good hotel
## 3 N A
## 4 The sauna is a wonderfull addition it s a shame I had to slalom in the cleaning staff halway almost filled with cleaning products Also the proximity to the cleaning staff s room added all the noise to what was supposed to be a quiet moment maybe it would be a good idea to relocate their general quarter
## 5 They have maintenance and I couldn t enjoy the balcony view
## 6 Reception not impressive building was under construction Room passage was way to small Room has also no space# Reorder columns in the specified order
ordered_data <- paris_hotels %>%
select(Hotel_Name, Average_Score, Positive_Review, Negative_Review)
# View the first few rows of the reordered data frame
head(ordered_data)## Hotel_Name Average_Score
## 1 Monhotel Lounge SPA 8.4
## 2 Monhotel Lounge SPA 8.4
## 3 Monhotel Lounge SPA 8.4
## 4 Monhotel Lounge SPA 8.4
## 5 Monhotel Lounge SPA 8.4
## 6 Monhotel Lounge SPA 8.4
## Positive_Review
## 1 Nice hotel Room was beautiful and bed very comfortable Did not expect big rooms in Paris so size was really ok with 2 windows Bathroom modern and new with a GREAT shower Staff was wonderfull since reservation They sent me an email prior to my arrival asking if they could help me in anyway and also asking if I had any special needs such as extra beds I had a last minute health problem in the family and had to arrive one day earlier and stay for 1 night less Audrey was wonderful and managed to receive me before and dealt herself with Booking to change my reservation I sent a note to her when I was entering my flight from Cannes to Paris and when I left the airplane she had already sorted everything out They even charged me a lower rate as on the day I arrived the room costed a little less In this same day Saturday June 25 I went out with friends and arrived really hungry at the hotel at 1 30AM and the guy from reception can t believe I forgot his name Maybe Hadesmi managed to open the restaurant and get to my room a nice aspargus risoto bread yougurt and a warm smile even thought riom service finishes at midnight Really had a very nice experience at Monhotel and recomend it
## 2 No Positive
## 3 Fantastic weekend with my partner We Would definitely stay here again Location was excellent and couldn t have asked for a better area to stay in Was so close to the metro and all the shops bars and restaurants Room was lovely and cosey and very modern with a classy edge to the facility
## 4 Good location really good breakfast maybe the best I ever had in Paris in a confortable and stylish setting Everything you need is provided on site you can literally arrive with your hands in your pockets Helpfull and attentive staff
## 5 I asked to be downgraded because I was paying for a balcony room while I couldn t enjoy the balcony Instead they upgraded me with the biggest room in the hotel for the same rate
## 6 The room has everything inside as a full package
## Negative_Review
## 1 No Negative
## 2 Not only did the staff on arrival ask to copy the details of my credit card this is OK but they also noted down the CVV number from the back of the credit card This means that the hotel has all the information to use my credit card for internet purchases Since this is unacceptable I asked the receptionist to delete the CVV but she refused This policy should be changed Otherwise good hotel
## 3 N A
## 4 The sauna is a wonderfull addition it s a shame I had to slalom in the cleaning staff halway almost filled with cleaning products Also the proximity to the cleaning staff s room added all the noise to what was supposed to be a quiet moment maybe it would be a good idea to relocate their general quarter
## 5 They have maintenance and I couldn t enjoy the balcony view
## 6 Reception not impressive building was under construction Room passage was way to small Room has also no spaceSummary stats
Basic statistical measures, such as mean and median, are calculated for the Average_Score variable. These measures help establish benchmarks for distinguishing high-rated and low-rated hotels, which will later be correlated with review keywords.
# Calculate the summary statistics for Average_Score
summary_stats <- summary(ordered_data$Average_Score)
print(summary_stats)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 6.800 8.100 8.500 8.409 8.800 9.800Visualization: Box Plot
To visualize the distribution of hotel ratings, a box plot is created. This visualization highlights the mean, median, and outliers, offering a clear understanding of rating trends across hotels in Paris.
# Create the boxplot
boxplot_graph <- ggplot(ordered_data, aes(x = "", y = Average_Score)) +
geom_boxplot(fill = "blue", color = "black") +
labs(title = "Boxplot of Average Hotel Ratings", y = "Average Score") +
theme_minimal()
boxplot_graph
ggsave("Boxplot_Average_Hotel_Ratings.png", boxplot_graph, width = 10, height = 6)Individual figures
Describe and show how you created the first figure. Why did you choose this figure type?
In showing the figures that you created, describe why you designed it the way you did. Why did you choose those colors, fonts, and other design elements? Does it convey truth?
Figure 1 : Core Keywords - Wordcloud & Bar plot
By analyzing positive and negative reviews of various hotels in Paris, frequently mentioned words are identified as core elements and visualized using a wordcloud for intuitive understanding. The size of each word corresponds to its frequency, making it effortless to pinpoint the most significant factors contributing to customer satisfaction and dissatisfaction. The approach allows hotels to easily prioritize areas for improvement while reinforcing their strengths to better meet customer expectations.
The keywords of each positive review and negative review are tokenized to extract the upper keyword after calculating the frequency. At this time, unnecessary or meaningless words are excluded and extracted again according to the analyst’s judgment. Repeat this until meaningful words are extracted.
Wordclouds were chosen for keyword analysis because they visually represent keyword frequency, with larger words indicating higher importance. This makes it easy to identify dominant themes in positive and negative reviews at a glance. By filtering out unnecessary words and focusing on meaningful ones, wordclouds provide a clear, intuitive summary, making them ideal for both exploratory analysis and presenting insights effectively.
Data processing for Wordcloud Visualization 1: Tokenize
# Extract key keywords from Positive_Reviews
# Tokenize words from positive reviews
p_token <- ordered_data %>%
unnest_tokens(word, Positive_Review)
# Remove stop words (common words with no meaningful context)
p_tidyToken <- p_token %>%
anti_join(stop_words)## Joining with `by = join_by(word)`# Count word frequencies and sort in descending order
p_tidyToken <- p_tidyToken %>%
count(word, sort = TRUE)
# Display the top 20 most frequent keywords
head(p_tidyToken, 20)## word n
## 1 location 26845
## 2 staff 24461
## 3 hotel 16355
## 4 friendly 10853
## 5 helpful 10514
## 6 nice 8710
## 7 clean 7847
## 8 breakfast 7452
## 9 comfortable 6815
## 10 excellent 6727
## 11 paris 5490
## 12 bed 5385
## 13 metro 5140
## 14 close 4324
## 15 positive 3876
## 16 perfect 3840
## 17 stay 3809
## 18 lovely 3196
## 19 quiet 3087
## 20 service 2883# Tokenize words from negative reviews
n_token <- ordered_data %>%
unnest_tokens(word, Negative_Review)
# Remove stop words (common words with no meaningful context)
n_tidyToken <- n_token %>%
anti_join(stop_words)## Joining with `by = join_by(word)`# Count word frequencies and sort in descending order
n_tidyToken <- n_tidyToken %>%
count(word, sort = TRUE)
# Display the top 20 most frequent keywords
head(n_tidyToken, 20)## word n
## 1 negative 16826
## 2 hotel 8715
## 3 breakfast 6723
## 4 staff 4054
## 5 bit 3099
## 6 bathroom 2920
## 7 didn 2476
## 8 shower 2400
## 9 bed 2385
## 10 night 2298
## 11 service 2094
## 12 time 1959
## 13 expensive 1852
## 14 stay 1847
## 15 paris 1828
## 16 day 1776
## 17 poor 1737
## 18 wifi 1724
## 19 price 1696
## 20 water 1693Data processing for Wordcloud Visualization 2: Data cleaning
#positive
p_tidyToken <- p_token %>%
anti_join(stop_words) %>% # Remove stop words
filter(!word %in% c("hotel", "paris", "positive", "nice", "comfortable", "excellent", "perfect", "stay", "lovely")) %>% # Exclude specified words
count(word, sort = TRUE) # Count word frequencies and sort in descending order## Joining with `by = join_by(word)`# Display the top 20 most frequent keywords
head(p_tidyToken, 20)## word n
## 1 location 26845
## 2 staff 24461
## 3 friendly 10853
## 4 helpful 10514
## 5 clean 7847
## 6 breakfast 7452
## 7 bed 5385
## 8 metro 5140
## 9 close 4324
## 10 quiet 3087
## 11 service 2883
## 12 station 2860
## 13 restaurants 2689
## 14 walk 2654
## 15 amazing 2548
## 16 view 2425
## 17 bathroom 2298
## 18 tower 2250
## 19 eiffel 2091
## 20 beautiful 1996# negative
n_tidyToken <- n_token %>%
anti_join(stop_words) %>% # Remove common stop words
filter(!word %in% c("negative", "hotel", "bit", "didn", "paris", "2", "4", "night", "time", "stay", "day", "poor")) %>% # Exclude specified words
count(word, sort = TRUE) # Count word frequencies and sort in descending order## Joining with `by = join_by(word)`# Display the top 20 most frequent keywords
head(n_tidyToken, 20)## word n
## 1 breakfast 6723
## 2 staff 4054
## 3 bathroom 2920
## 4 shower 2400
## 5 bed 2385
## 6 service 2094
## 7 expensive 1852
## 8 wifi 1724
## 9 price 1696
## 10 water 1693
## 11 coffee 1639
## 12 reception 1551
## 13 floor 1521
## 14 bar 1506
## 15 check 1434
## 16 bad 1414
## 17 location 1313
## 18 wasn 1304
## 19 door 1302
## 20 air 1184Visualization of Wordclouds
The Wordcloud visualization highlights the most frequently mentioned keywords from positive and negative reviews, providing a clear view of customer priorities and concerns.
Positive reviews keywords Positive reviews often feature words like “clean,” “friendly,” and “location,” indicating that cleanliness, service quality, and accessibility are primary drivers of customer satisfaction in Parisian hotels.
Negative reviews keywords negative reviews frequently include terms such as “breakfast,” “price,” and “noise,” revealing common pain points.
Word cloud for positive keywords
# Open a PNG device to save the wordcloud
png("Wordcloud_Positive_Keywords.png", width = 800, height = 600)
# Activate the graphics device and generate the wordcloud
wordcloud(
words = p_tidyToken$word,
freq = p_tidyToken$n,
min.freq = 1000,
max.words = 100,
random.order = FALSE,
rot.per = 0.35,
colors = brewer.pal(8, "Set1")
)
# Close the graphics device to save the image
dev.off()## quartz_off_screen
## 2# Activate the graphics device and generate the wordcloud
wordcloud(
words = p_tidyToken$word,
freq = p_tidyToken$n,
min.freq = 1000,
max.words = 100,
random.order = FALSE,
rot.per = 0.35,
colors = brewer.pal(8, "Set1")
)
Word cloud for nagative keywords
# Open a PNG device to save the word cloud
png("Wordcloud_Negative_Keywords.png", width = 800, height = 600)
# Generate the word cloud
wordcloud(
words = n_tidyToken$word,
freq = n_tidyToken$n,
min.freq = 1000,
max.words = 100,
random.order = FALSE,
rot.per = 0.35,
colors = brewer.pal(8, "Set1")
)
# Close the graphics device to save the image
dev.off()## quartz_off_screen
## 2# Generate the word cloud
wordcloud(
words = n_tidyToken$word,
freq = n_tidyToken$n,
min.freq = 1000,
max.words = 100,
random.order = FALSE,
rot.per = 0.35,
colors = brewer.pal(8, "Set1")
)
Figure 2: Bar Plot - Keyword Frequency Analysis
While word clouds provide an engaging, high-level overview of text data by visually emphasizing frequently mentioned keywords, they lack the precision needed for detailed analysis. To address this limitation, bar graphs were used as a complementary tool to provide a more analytical and structured approach.
Bar graphs effectively display frequency distributions, allowing for clear and precise comparisons between keywords. This approach helps prioritize the most impactful factors by showcasing exact values and trends, making them an ideal choice for gaining deeper insights into the patterns present in both positive and negative review content.
This figure visualizes the most frequently mentioned positive and negative keywords from customer reviews in high-rated and low-rated hotels, respectively. The primary objective is to identify the key factors contributing to customer satisfaction and dissatisfaction, based on review content.
Data processing for Bar plot visualization
# Predefined positive/negative keywords from reviews
positive_keywords <- c("location", "staff", "friendly", "helpful", "clean", "breakfast", "bed", "metro", "quiet", "service")
negative_keywords <- c("breakfast", "staff", "bathroom", "shower", "bed",
"service", "expensive", "wifi", "price", "water")
# Setting rating criteria (High-rating vs Low-rating hotels)
high_rating_hotels <- ordered_data %>%
filter(Average_Score > median(ordered_data$Average_Score))
low_rating_hotels <- ordered_data %>%
filter(Average_Score <= median(ordered_data$Average_Score))
# Analyze the frequency of positive keywords in high-rating hotels
high_positive_token <- high_rating_hotels %>%
unnest_tokens(word, Positive_Review) %>%
filter(word %in% positive_keywords) %>% # Filter predefined positive keywords from reviews
anti_join(stop_words) %>%
count(word, sort = TRUE)## Joining with `by = join_by(word)`# Analyze the frequency of negative keywords in low-rating hotels
low_negative_token <- low_rating_hotels %>%
unnest_tokens(word, Negative_Review) %>%
filter(word %in% negative_keywords) %>% # Filter predefined negative keywords from reviews
anti_join(stop_words) %>%
count(word, sort = TRUE)## Joining with `by = join_by(word)`Bar Plot Visualization - keywords
The analysis reveals that excellent service, cleanliness, and convenient locations are the primary drivers of customer satisfaction in high-rated hotels, as reflected in positive keywords like staff, location, and clean. On the other hand, dissatisfaction in low-rated hotels stems from issues such as poor breakfast quality, inadequate bathroom conditions, and perceived high costs, highlighted by keywords like breakfast, bathroom, and expensive.
To enhance guest experiences, high-rated hotels should continue prioritizing their strengths while addressing secondary concerns like breakfast quality. Conversely, low-rated hotels must focus on improving basic amenities and staff service to resolve key pain points and boost overall satisfaction.
Positive keywords from review data: The bar graph for positive keywords highlights terms like “location,” “staff,” and “clean,” which frequently appear in high-rated hotel reviews. These keywords suggest that convenient locations, excellent staff service, and cleanliness are crucial for achieving high customer satisfaction. Hotels with high ratings often prioritize cleanliness, service quality, and prime locations. These factors are consistently valued by customers and directly influence satisfaction levels.
Negative keywords from review data: The bar graph for negative keywords shows terms like “breakfast,” “bathroom,” and “expensive,” which are commonly mentioned in low-rated hotel reviews. These keywords indicate areas where hotels may underperform, leading to dissatisfaction among guests. Guests at low-rated hotels often express dissatisfaction with breakfast quality, bathroom conditions, and perceived high costs. Addressing these areas could help improve customer satisfaction and overall ratings.
Keywords frequency from Positive reviews in High-Rating Hotels
bar_chart_high_positive <- ggplot(high_positive_token, aes(x = reorder(word, n), y = n)) +
geom_col(fill = "steelblue") + # Single color for simplicity
labs(title = "Keyword Frequency in High-Rating Hotels (Positive Reviews)",
x = "Keyword",
y = "Frequency") +
theme_minimal() +
coord_flip() # Flip the Y-axis for better readability
bar_chart_high_positive
# Save the bar chart as an image
ggsave("Keyword_Frequency_High_Rating_Hotels.png", bar_chart_high_positive, width = 10, height = 6)Keywords frequency from Negative revies in Low-Rating Hotels
bar_chart_low_negative <- ggplot(low_negative_token, aes(x = reorder(word, n), y = n)) +
geom_col(fill = "red") + # Single color for simplicity
labs(title = "Keyword Frequency in Low-Rating Hotels (Negative Reviews)",
x = "Keyword",
y = "Frequency") +
theme_minimal() +
coord_flip() # Flip the Y-axis for better readability
bar_chart_low_negative
# Save the bar chart as an image
ggsave("Keyword_Frequency_Low_Rating_Hotels.png", bar_chart_low_negative, width = 10, height = 6)Figure 3: Bar Plot - Sentiment Analysis
To complement keyword analysis from another perspective, sentiment analysis provides a more nuanced understanding of guest feedback by categorizing it into distinct positive and negative sentiments. Unlike simple keyword frequency visualization, sentiment analysis captures the emotional tone of reviews, offering deeper insights into the factors that drive satisfaction or dissatisfaction. By emphasizing the polarity of sentiments, this approach allows hotels to identify not only the most frequently mentioned aspects but also the emotional weight behind them. This makes sentiment analysis a powerful tool for prioritizing actionable improvements and strategically enhancing guest experiences.
Data processing for Bar plot visualization
# Attach sentiment lexicon for analysis (using bing)
bing_sentiments <- get_sentiments("bing")
# Analyze sentiment in Positive_Review
positive_sentiment <- ordered_data %>%
unnest_tokens(word, Positive_Review) %>%
inner_join(bing_sentiments, by = "word") %>%
count(sentiment, sort = TRUE) %>%
mutate(review_type = "Positive")
# Analyze sentiment in Negative_Review
negative_sentiment <- ordered_data %>%
unnest_tokens(word, Negative_Review) %>%
inner_join(bing_sentiments, by = "word") %>%
count(sentiment, sort = TRUE) %>%
mutate(review_type = "Negative")
# Combine both datasets
combined_sentiments <- bind_rows(positive_sentiment, negative_sentiment)Bar Plot Visualization - sentiment comparison
The bar plot highlights the frequency of positive and negative sentiments in customer reviews. Positive sentiments, represented by words like clean and friendly, dominate in positive reviews, while negative sentiments, including expensive and uncomfortable, are prominent in negative reviews.
This figure demonstrates the importance of enhancing positive elements such as cleanliness and staff service while addressing recurring issues related to pricing and amenities. The sentiment comparison provides a balanced view of customer feedback, helping hotels prioritize efforts to enhance guest satisfaction effectively.
Positive Sentiments: Keywords like staff, cleanliness, and friendly frequently appear in positive reviews, reinforcing their role in customer satisfaction. These aspects are consistently associated with high ratings and positive guest experiences.
Negative Sentiments: Keywords like breakfast, price, and uncomfortable dominate negative reviews, indicating areas where hotels underperform and need improvement.
bar plot Visualization
sentiment_comparison_chart <- ggplot(combined_sentiments, aes(x = sentiment, y = n, fill = review_type)) +
geom_bar(stat = "identity", position = "dodge") +
labs(
title = "Sentiment Comparison for Positive and Negative Reviews",
x = "Sentiment",
y = "Frequency",
fill = "Review Type"
) +
theme_minimal()
sentiment_comparison_chart
# Save the bar chart as an image
ggsave("Sentiment_Comparison_Positive_Negative_Reviews.png", sentiment_comparison_chart, width = 10, height = 6)Figure 4: Scattor plot with Trend Line - Keyword Ratios vs. Ratings
This figure visualizes the relationship between keyword ratios (both from positive and negative reviews) and hotel ratings. By plotting keyword ratios against the average ratings of hotels, we aim to uncover patterns that indicate how strongly the presence of specific keywords in reviews correlates with customer satisfaction.
The scatter plot was chosen because it effectively displays correlations and distributions of data points, enabling us to analyze trends and relationships between variables. The addition of a trend line highlights overarching patterns, such as whether positive keyword ratios are directly proportional to higher ratings or whether negative keyword ratios are inversely related to ratings. This design ensures the visualization is both intuitive and data-driven.
Data processing for Scattor plot visualization 1
# Define keywords from positive and negative reviews
positive_keywords <- c("location", "staff", "friendly", "helpful", "clean",
"breakfast", "bed", "metro", "quiet", "service")
negative_keywords <- c("breakfast", "staff", "bathroom", "shower", "bed",
"service", "expensive", "wifi", "price", "water")
# Define a unified list of keywords
all_keywords <- c("location", "staff", "breakfast", "service", "clean",
"bed", "metro", "bathroom", "close")
# Group data by hotel and prepare the initial summary
hotel_keywords_summary <- ordered_data %>%
group_by(Hotel_Name) %>%
summarise(
# Calculate the average score
average_score = mean(Average_Score, na.rm = TRUE),
# Count the number of positive reviews
num_positive_reviews = sum(!is.na(Positive_Review)),
# Count the number of negative reviews
num_negative_reviews = sum(!is.na(Negative_Review)),
# Calculate the total frequency of positive keywords in positive reviews
p_keywords_num = sum(str_count(tolower(Positive_Review), paste(positive_keywords, collapse = "|"))),
# Calculate the total frequency of negative keywords in negative reviews
n_keywords_num = sum(str_count(tolower(Negative_Review), paste(negative_keywords, collapse = "|")))
) %>%
ungroup() %>%
# Calculate positive and negative keyword ratios
mutate(
positive_keyword_ratio = p_keywords_num / num_positive_reviews,
negative_keyword_ratio = n_keywords_num / num_negative_reviews
)
# Use a for loop to calculate the frequency of each keyword in positive/negative reviews and convert to ratios
for (keyword in all_keywords) {
# Calculate the frequency of the keyword in positive reviews and convert to ratio
hotel_keywords_summary <- hotel_keywords_summary %>%
mutate(!!paste0("p_", keyword, "_ratio") :=
ordered_data %>%
group_by(Hotel_Name) %>%
summarise(keyword_count = sum(str_count(tolower(Positive_Review), fixed(keyword)), na.rm = TRUE)) %>%
pull(keyword_count) / num_positive_reviews)
# Calculate the frequency of the keyword in negative reviews and convert to ratio
hotel_keywords_summary <- hotel_keywords_summary %>%
mutate(!!paste0("n_", keyword, "_ratio") :=
ordered_data %>%
group_by(Hotel_Name) %>%
summarise(keyword_count = sum(str_count(tolower(Negative_Review), fixed(keyword)), na.rm = TRUE)) %>%
pull(keyword_count) / num_negative_reviews)
}
# View the full dataset sorted by average score
hotel_keywords_summary <- hotel_keywords_summary %>%
arrange(desc(average_score))
head(hotel_keywords_summary, 1000)## # A tibble: 458 × 26
## Hotel_Name average_score num_positive_reviews num_negative_reviews
## <chr> <dbl> <int> <int>
## 1 Ritz Paris 9.8 28 28
## 2 H tel de La Tamise E… 9.6 61 61
## 3 Hotel The Peninsula … 9.5 58 58
## 4 Le Narcisse Blanc Spa 9.5 57 57
## 5 Goralska R sidences … 9.4 24 24
## 6 H tel D Aubusson 9.4 294 294
## 7 Hotel Eiffel Blomet 9.4 15 15
## 8 Hotel Monge 9.4 115 115
## 9 La Chambre du Marais 9.4 88 88
## 10 Nolinski Paris 9.4 113 113
## # ℹ 448 more rows
## # ℹ 22 more variables: p_keywords_num <int>, n_keywords_num <int>,
## # positive_keyword_ratio <dbl>, negative_keyword_ratio <dbl>,
## # p_location_ratio <dbl>, n_location_ratio <dbl>, p_staff_ratio <dbl>,
## # n_staff_ratio <dbl>, p_breakfast_ratio <dbl>, n_breakfast_ratio <dbl>,
## # p_service_ratio <dbl>, n_service_ratio <dbl>, p_clean_ratio <dbl>,
## # n_clean_ratio <dbl>, p_bed_ratio <dbl>, n_bed_ratio <dbl>, …# Create and view a smaller dataset with key columns only
visible_columns <- c("Hotel_Name", "average_score", "positive_keyword_ratio", "negative_keyword_ratio")
hidden_columns <- setdiff(names(hotel_keywords_summary), visible_columns)
hotel_keywords_summary_visible <- hotel_keywords_summary %>%
select(all_of(visible_columns))
head(hotel_keywords_summary_visible, 1000)## # A tibble: 458 × 4
## Hotel_Name average_score positive_keyword_ratio negative_keyword_ratio
## <chr> <dbl> <dbl> <dbl>
## 1 Ritz Paris 9.8 1.04 0.25
## 2 H tel de La Tami… 9.6 2.46 0.148
## 3 Hotel The Penins… 9.5 1.24 0.328
## 4 Le Narcisse Blan… 9.5 1.81 0.667
## 5 Goralska R siden… 9.4 1.79 0.542
## 6 H tel D Aubusson 9.4 1.90 0.340
## 7 Hotel Eiffel Blo… 9.4 2.8 0.4
## 8 Hotel Monge 9.4 2.35 0.287
## 9 La Chambre du Ma… 9.4 2.25 0.216
## 10 Nolinski Paris 9.4 1.53 0.310
## # ℹ 448 more rowsScattor Plot Visualization - Positve vs. Negative
The scatter plot shows the correlation between average hotel ratings (x-axis) and the frequency ratios of positive (blue points) and negative (red points) keywords (y-axis) in customer reviews. The trend lines for keywords from positive and negative reviews highlight the overall patterns, offering a balanced perspective on the factors driving satisfaction and dissatisfaction.
This data is calculated and applied in a ratio rather than the frequency. Since the number of reviews for each hotel is different, the total number is divided by the number of reviews for the hotel for accurate judgment.
By visualizing the relationship between keyword ratios and ratings, this analysis provides actionable insights for hotels to strategically enhance their services. Leveraging strengths like cleanliness and service quality while prioritizing improvements in key pain points, such as pricing and amenities, can help hotels achieve sustainable growth and greater customer satisfaction.
Keywords from Positive reviews (Blue Points): The upward-sloping trend line for positive keywords illustrates a clear positive correlation between the frequency of positive terms in reviews and higher hotel ratings. At this time, the key drivers are the 10 keywords previously obtained through the word cloud: [location, staff, friendly, helpful, clean, breast, bed, metro, quiet, service] Hotels achieving higher ratings often receive reviews highlighting aspects such as staff, clean, and location, underscoring the importance of these factors in guest satisfaction. And keyword staff, service, bed, bathroom and breakfast shows the steeper the upward curve in positive review.
Keywords from Negative reviews (Red Points): The downward-sloping trend line for negative keywords indicates a negative correlation between the frequency of negative terms and hotel ratings.At this time, the key drivers are the 10 keywords previously obtained through the word cloud: [breakfast, staff, bathroom, show, bed, service, expansive, wifi, price, water] Lower-rated hotels frequently receive reviews with terms such as breakfast, bathroom, and expensive, pointing to critical areas where improvements are needed to mitigate dissatisfaction and boost ratings. And keyword staff, service, bed, bathroom and breakfast shows the steeper the downward curve in negative reviews
# Visualizing the relationship between ratings and the ratio of positive/negative keywords
rating_keyword_ratio_plot <- ggplot(hotel_keywords_summary, aes(x = average_score)) +
# Scatter plot for the ratio of positive keywords
geom_point(aes(y = positive_keyword_ratio, color = "Positive Reviews"), alpha = 0.6) +
geom_smooth(aes(y = positive_keyword_ratio, color = "Positive Reviews"),
method = "lm", se = FALSE, linetype = "solid") +
# Scatter plot for the ratio of negative keywords
geom_point(aes(y = negative_keyword_ratio, color = "Negative Reviews"), alpha = 0.6) +
geom_smooth(aes(y = negative_keyword_ratio, color = "Negative Reviews"),
method = "lm", se = FALSE, linetype = "solid") +
# Title and axis labels
labs(
title = "Correlation Between Rating and Keyword Frequency Ratio",
x = "Average Score",
y = "Keyword Frequency Ratio",
subtitle = "Blue points: Positive Keywords Ratio, Red points: Negative Keywords Ratio"
) +
# Color legend
scale_color_manual(values = c("Positive Reviews" = "blue", "Negative Reviews" = "red")) +
# Minimal theme for clean visuals
theme_minimal()
rating_keyword_ratio_plot## `geom_smooth()` using formula = 'y ~ x'
## `geom_smooth()` using formula = 'y ~ x'
# Save the scatter plot as an image
ggsave("Rating_Keyword_Frequency_Ratio.png", rating_keyword_ratio_plot, width = 12, height = 8)## `geom_smooth()` using formula = 'y ~ x'
## `geom_smooth()` using formula = 'y ~ x'Data processing for Facet scattor plot visualization 2
# Select relevant columns and transform the data
facet_data <- hotel_keywords_summary %>%
pivot_longer(
cols = matches("^(p_|n_).+_ratio$"), # Select columns starting with p_ or n_ and ending with _ratio
names_to = "keyword_type",
values_to = "ratio"
) %>%
mutate(
sentiment = ifelse(grepl("^p_", keyword_type), "Positive", "Negative"), # Separate Positive/Negative sentiments
keyword = gsub("^(p_|n_)", "", keyword_type) %>% gsub("_ratio$", "", .) # Remove p_/n_ and _ratio from column names
)
# Filter data: Use only keywords included in all_keywords
facet_data <- facet_data %>%
filter(keyword %in% all_keywords)
# Check the data structure and preview
str(facet_data)## tibble [8,244 × 12] (S3: tbl_df/tbl/data.frame)
## $ Hotel_Name : chr [1:8244] "Ritz Paris" "Ritz Paris" "Ritz Paris" "Ritz Paris" ...
## $ average_score : num [1:8244] 9.8 9.8 9.8 9.8 9.8 9.8 9.8 9.8 9.8 9.8 ...
## $ num_positive_reviews : int [1:8244] 28 28 28 28 28 28 28 28 28 28 ...
## $ num_negative_reviews : int [1:8244] 28 28 28 28 28 28 28 28 28 28 ...
## $ p_keywords_num : int [1:8244] 29 29 29 29 29 29 29 29 29 29 ...
## $ n_keywords_num : int [1:8244] 7 7 7 7 7 7 7 7 7 7 ...
## $ positive_keyword_ratio: num [1:8244] 1.04 1.04 1.04 1.04 1.04 ...
## $ negative_keyword_ratio: num [1:8244] 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 ...
## $ keyword_type : chr [1:8244] "p_location_ratio" "n_location_ratio" "p_staff_ratio" "n_staff_ratio" ...
## $ ratio : num [1:8244] 0.0357 0 0.4286 0.0357 0.0714 ...
## $ sentiment : chr [1:8244] "Positive" "Negative" "Positive" "Negative" ...
## $ keyword : chr [1:8244] "location" "location" "staff" "staff" ...head(facet_data)## # A tibble: 6 × 12
## Hotel_Name average_score num_positive_reviews num_negative_reviews
## <chr> <dbl> <int> <int>
## 1 Ritz Paris 9.8 28 28
## 2 Ritz Paris 9.8 28 28
## 3 Ritz Paris 9.8 28 28
## 4 Ritz Paris 9.8 28 28
## 5 Ritz Paris 9.8 28 28
## 6 Ritz Paris 9.8 28 28
## # ℹ 8 more variables: p_keywords_num <int>, n_keywords_num <int>,
## # positive_keyword_ratio <dbl>, negative_keyword_ratio <dbl>,
## # keyword_type <chr>, ratio <dbl>, sentiment <chr>, keyword <chr>Facet scattor Plot Visualization - Keywords (Positive vs. Negative)
The earlier scatter plot provided a general overview of the correlation between positive and negative keyword ratios and hotel ratings, highlighting collective trends. However, this aggregated view lacked the granularity required to evaluate the individual impact of specific keywords. To address this limitation, a facet scatter plot was created to break down the relationships between specific keywords (e.g., bathroom, bed, breakfast, etc.) and average hotel ratings. By visualizing the keyword ratios in positive (blue trend lines) and negative (red trend lines) reviews across individual panels, the facet plot allows for a more detailed and targeted analysis of satisfaction and dissatisfaction drivers.
Facet scatter plots were specifically chosen because they enable side-by-side comparisons of multiple variables, helping to uncover nuanced patterns for each keyword. This design ensures that each keyword’s unique contribution to ratings is clear and actionable. For instance, keywords like clean, staff, and service exhibit strong positive trends, indicating their frequent mentions in positive reviews are closely tied to higher ratings. Conversely, keywords like breakfast and bathroom display flat or slightly negative trends in negative reviews, reinforcing their association with lower ratings. This granular analysis enables hotels to focus on addressing specific weaknesses while maintaining key strengths to improve overall guest satisfaction and ratings.
Positive Keywords (Blue Trend Lines): Keywords like clean, staff, and service exhibit a strong positive trend, indicating that frequent mentions of these terms in positive reviews are strongly associated with higher ratings. Other keywords like location and bed show a moderate positive trend, suggesting they contribute positively but less strongly compared to clean and staff.
Negative Keywords (Red Trend Lines): Keywords like bathroom and breakfast show a flat or slightly negative trend, highlighting that their frequent mentions in negative reviews are associated with lower ratings. The red lines for most keywords remain flat or negatively sloped, reinforcing the negative impact of these aspects on customer satisfaction.
# Create Facet Scatter Plot
facet_scatter_plot <- ggplot(facet_data, aes(x = average_score, y = ratio, color = sentiment)) +
geom_point(alpha = 0.15) +
geom_smooth(method = "lm", se = FALSE, linetype = "solid", alpha = 0.7) +
facet_wrap(~ keyword, scales = "free_y") + # Create facets for each keyword
labs(
title = "Keyword Ratio vs Average Score by Sentiment",
x = "Average Score",
y = "Keyword Ratio",
color = "Reviews"
) +
scale_color_manual(values = c("Positive" = "blue", "Negative" = "red")) +
theme_minimal() +
theme(
strip.text = element_text(size = 10, face = "bold"),
legend.position = "top"
)
facet_scatter_plot## `geom_smooth()` using formula = 'y ~ x'
# Save the Facet Scatter Plot as an image
ggsave("Facet_Scatter_Keyword_Ratio.png", facet_scatter_plot, width = 14, height = 10)## `geom_smooth()` using formula = 'y ~ x'Conclusion
The analysis of customer reviews using various visualizations highlights the key factors influencing hotel ratings and provides actionable insights for improving guest satisfaction. Across all figures, common themes such as cleanliness, staff service, location, and pricing emerge as the most significant contributors to customer experiences.
Key Drivers of Satisfaction Positive aspects such as cleanliness, staff service, and location consistently emerge as the most influential factors in achieving high customer satisfaction. Hotels with higher ratings frequently receive positive mentions of these aspects, underscoring their importance in delivering exceptional guest experiences. Keywords like friendly and helpful further highlight the value of personalized and attentive service in fostering positive impressions. Maintaining and enhancing these strengths is crucial for high-rated hotels to sustain their competitive edge.
Key Areas of Improvement Recurring issues like breakfast quality, bathroom conditions, and pricing concerns are prominent in negative reviews and correlate with lower ratings. These pain points indicate where low-rated hotels commonly underperform, leading to dissatisfaction among guests. Addressing these weaknesses by improving basic amenities, such as breakfast offerings and bathroom maintenance, and providing better value for money is essential for boosting guest satisfaction and overall ratings.
Strategic Implications The scatter plot analysis revealed strong correlations between higher positive keyword ratios and higher ratings, as well as between negative keyword ratios and lower ratings. This indicates that fostering positive guest experiences through excellent service and cleanliness while minimizing negative experiences related to pricing and amenities can significantly impact customer perceptions. Furthermore, the detailed breakdown of individual keywords demonstrates the need for targeted interventions to address specific areas of dissatisfaction, such as breakfast and bathroom, while reinforcing strengths like staff and cleanliness.
Actionable Insights For High-Rated Hotels: Continue prioritizing key strengths, such as cleanliness and staff service, while addressing secondary concerns like breakfast quality to maintain and improve guest satisfaction. For Low-Rated Hotels: Focus on resolving operational issues related to pricing, breakfast offerings, and bathroom amenities. Investing in these areas can mitigate dissatisfaction and elevate overall ratings. Holistic Perspective
The sentiment analysis reinforces these findings by providing a balanced view of customer feedback. Positive sentiments dominate high-rated reviews, reflecting the satisfaction derived from clean facilities and friendly staff. Conversely, negative sentiments in low-rated reviews highlight dissatisfaction with price and comfort-related issues. By leveraging these insights, hotels can strategically align their resources to enhance strengths, address weaknesses, and create superior guest experiences.
In conclusion, this analysis provides a roadmap for hoteliers to optimize their services and prioritize operational improvements effectively. By focusing on both strengths and areas for improvement, hotels can drive customer satisfaction, improve their competitive positioning, and achieve sustainable success in the hospitality industry.