1. (c)C
opyrighted
- 0 -
Final Year Project 2013/2014
Topic: Online Hotel Reviews
Extraction and Analysis
Report
Student Name : LIM Yin Yun, Eleanor
Student ID : 10701230D
Programme : BSc (HONS) in Enterprise Information System
Programme Code : 61031 – FIS
Supervisor : Dr. LIU Nga Kwok, James
Co-Examiner : Dr. YIU, Ken
Moderator : Dr. LEUNG, Hareton
Submission Date : 14th April 2014
2. (c)C
opyrighted
- 1 -
Abstract
Nowadays, the accessibility to the Internet has improved significantly. Travelers are
switching from conventional media to the Internet for travel planning. Online hotel
booking is perceived as a high risk purchase that involves a lot of uncertainties.
Therefore, travelers sought for means that could reduce the risk and aid them in
making sound judgments. Travelers believe online reviews could help them as these
reviews are written voluntarily by other travelers and they are more consumer-
oriented.
From a business perspective, the vast amount of online reviews can be utilized by
the company to create competitive advantage. As these online reviews contain lots
of information about the preferences of customers, if being used strategically, the
company would gain insights and valuable information. The importance and
utilization of big data can differentiate one from its fellow rivals.
This study extracted online hotel reviews written in English from one of the
renowned travel website, www.hostelworld.com with a focus on the City of Light,
Paris by using a web crawler known as WebHarvy. WebHarvy is chosen due to its
simplicity and efficiency in data crawling. After removing redundant data, a total of
4500 online hotel reviews during the time period September 2009 – January 2014 is
used to conduct quantitative analysis. Each review contains rating of the hotel given
by the author, profile of the author, date of posting and textual comment.
The raw data is then pre-processed by removing stop words based on a stop word
list generated by the Information Retrieval Group of University of Glasgow. A
sentiment corpus provided by previous study with a focus in the hotel industry is
used and further expanded by implementing Rita WordNet, which is a Java API
developed by Daniel Howe. Sentiment analysis is done according to a hotel feature
list which was built based on several recent studies, including facilities, location,
room, security, service, and value.
3. (c)C
opyrighted
- 2 -
The sentiment is judged by counting and comparing the number of positive and
negative words in each sentence. If the number of positive words is greater than
negative words, then the sentence is categorized as a positive review, and vice versa.
However, if the number of positive words is equal to negative words, then the
sentence is categorized as a neutral review. The sentimental results are then
updated into the excel file.
Observations and inferences are made to find out the overall performance of
Parisian hotels, travelers’ preferences in choosing hotels and online review writing
behaviors. Several demographic comparisons are done, including continents, age
groups and genders.
Some of the interesting results include:
- Parisian hotels are performing well in terms of value, service and location.
However, there are rooms for improvements in terms of facilities and room.
- European travelers are most concerned about hotel room.
- Oceania travelers are considerably concerned about hotel value.
- Asian travelers are least concerned about hotel service.
- Hotel security bears little importance to travelers in general.
- The biggest group of travelers to Paris is European.
- Summer is the peak season to travel to Paris.
- Females are traveling more than ever and dominating the online review
community.
- The initial impulse of writing an online review is due to positive travel
experience.
- A significant amount of travelers remained anonymous in writing online
reviews.
This study is concluded by giving recommendations to hotel managers on further
improvement in hotels
4. (c)C
opyrighted
- 3 -
Acknowledgement
It has never been easy throughout the final year. There were problems and
challenges that hit on me along the way. But I am glad it has now come to an end and
I would like to take this opportunity to express my gratitude to all the people and
parties who have been supported me throughout this tough period.
First, I would like to express my gratitude to Dr. James Liu for his professional
guidance and support throughout the year. Dr. James Liu is a dedicated teacher who
is willing to go beyond his limits by spending his valuable time discussing my
project and ensuring I am on the right track. His advices are beneficial and have
provided me with provocative thoughts. In addition, I would like to thank my tutor,
Remy Hu for his advices and encouragement along the way.
I would also like to thank my co-examiner, Dr. Ken Yiu and my moderator, Dr.
Hareton Leung for dedicating their time to attend my presentation and commenting
on my project.
Last but not least, I would like to thank my family and friends whom have always
been there to support me during this though period, especially Fabian who has
always been there and showed constant support. Without you, I would never make
it.
5. (c)C
opyrighted
- 4 -
Contents
Abstract................................................................................................................................................................... - 1 -
Acknowledgement.............................................................................................................................................. - 3 -
1. Introduction................................................................................................................................................ - 9 -
2. Problem Statement.................................................................................................................................- 11 -
2.1 Insights from consumers...............................................................................................................- 11 -
2.2 Phenomenon of electronic word of mouth (eWOM)..........................................................- 11 -
2.3 Nature of products/services ........................................................................................................- 11 -
3. Literature Review...................................................................................................................................- 12 -
3.1 Online Travel Reviews....................................................................................................................- 12 -
3.2 Research context: Hostelworld.com .........................................................................................- 13 -
3.2.1 Review at Hostelworld.com..............................................................................................- 13 -
3.3 Natural Language Processing ......................................................................................................- 15 -
4. Project Methodology..............................................................................................................................- 16 -
4.1 Project Schedule................................................................................................................................- 16 -
4.2 Project Flow ........................................................................................................................................- 17 -
4.3 Data Collection...................................................................................................................................- 18 -
4.3.1 Sampling Data.........................................................................................................................- 20 -
4.4 Data Pre-Processing.........................................................................................................................- 21 -
4.5 Data Processing .................................................................................................................................- 22 -
4.5.1 Hotel feature list generation.............................................................................................- 22 -
4.5.2 Hotel feature matching.......................................................................................................- 22 -
4.5.3 Hotel feature and review sentence sentiment analysis ........................................- 22 -
4.6 Data Post-Processing.......................................................................................................................- 24 -
5 Data Analysis.............................................................................................................................................- 25 -
5.1 The Performance of Various Parisian Hotel Features .......................................................- 25 -
5.2 The Interest of Various Hotel Features among Different Continents..........................- 27 -
5.2.1 Hotel Facilities........................................................................................................................- 27 -
5.2.2 Hotel Location ........................................................................................................................- 28 -
5.2.3 Hotel Room..............................................................................................................................- 29 -
5.2.4 Hotel Security.........................................................................................................................- 30 -
5.2.5 Hotel Service...........................................................................................................................- 31 -
5.2.6 Hotel Value...............................................................................................................................- 32 -
6. (c)C
opyrighted
- 5 -
5.3 The Interest of Various Hotel Features among Different Age Groups........................- 33 -
5.3.1 Hotel Facilities........................................................................................................................- 33 -
5.3.2 Hotel Location ........................................................................................................................- 34 -
5.3.3 Hotel Room..............................................................................................................................- 35 -
5.3.4 Hotel Security.........................................................................................................................- 36 -
5.3.5 Hotel Service...........................................................................................................................- 37 -
5.3.6 Hotel Value...............................................................................................................................- 38 -
5.4 The Interest of Various Hotel Features among Different Genders...............................- 39 -
5.4.1 Hotel Facilities........................................................................................................................- 39 -
5.4.2 Hotel Location ........................................................................................................................- 40 -
5.4.3 Hotel Room..............................................................................................................................- 41 -
5.4.4 Hotel Security.........................................................................................................................- 42 -
5.4.5 Hotel Service...........................................................................................................................- 43 -
5.4.6 Hotel Value...............................................................................................................................- 44 -
5.5 Travel Frequency to Paris.............................................................................................................- 45 -
5.5.1 Among Different Continents.............................................................................................- 45 -
5.5.2 Among Different Age Groups ...........................................................................................- 47 -
5.5.3 Among Different Genders..................................................................................................- 49 -
5.6 Travel Seasons to Paris...................................................................................................................- 50 -
5.7 Online Review Writing....................................................................................................................- 51 -
5.7.1 Participation in Online Review Writing Between Male and Female................- 51 -
5.7.2 Participation in Online Review Writing Among Different Authors..................- 52 -
5.7.3 The Initial Impulse of Online Review Writing...........................................................- 53 -
5.7.4 Positive Online Review Writing Between Male and Female...............................- 54 -
5.7.5 Negative Online Review Writing Between Male and Female..............................- 55 -
6 Limitations.................................................................................................................................................- 56 -
6.1 Limited domain and sentiment corpus .......................................................................................- 56 -
6.2 Limited data set.....................................................................................................................................- 56 -
6.3 Unilingual.................................................................................................................................................- 56 -
7 Future Work..............................................................................................................................................- 57 -
7.1 Develop more complicated sentiment algorithm................................................................- 57 -
7.2 Expand domain of study ................................................................................................................- 57 -
7.3 Improve method of data analysis...............................................................................................- 57 -
7. (c)C
opyrighted
- 6 -
7.4 Consider Chinese review websites and data set..................................................................- 57 -
8 Recommendations..................................................................................................................................- 58 -
8.1 Improve hotel facilities and room..............................................................................................- 58 -
8.2 Focus on travelers between ages 18-30..................................................................................- 58 -
8.3 Attract Chinese travelers...............................................................................................................- 58 -
8.4 Consider negative reviews too....................................................................................................- 58 -
9 Conclusion..................................................................................................................................................- 59 -
10 References.............................................................................................................................................- 60 -
11 Appendix................................................................................................................................................- 66 -
11.1 Stop Words Removal.......................................................................................................................- 66 -
11.2 Hotel Feature Matching..................................................................................................................- 67 -
11.3 Sentiment Analysis...........................................................................................................................- 68 -
8. (c)C
opyrighted
- 7 -
List of Figures
Figure 1-a International tourist arrivals (UNWTO, 2013)----------------------------------------- 9 -
Figure 1-b Global top 20 destination cities (Master Card Global Destination Cities Index,
2013)-------------------------------------------------------------------------------------------------------- 9 -
Figure 1-c Iconic landmarks of Paris --------------------------------------------------------------- - 10 -
Figure 3-a Homepage of Hostelworld.com (Hostelworld.com, 2013)------------------------ - 13 -
Figure 3-b Reviews at Hostelworld.com (Hostelworld.com, 2013) -------------------------- - 14 -
Figure 4-a Project schedule -------------------------------------------------------------------------- - 16 -
Figure 4-b Project flow ------------------------------------------------------------------------------- - 17 -
Figure 4-c Screen shot of WebHarvy --------------------------------------------------------------- - 18 -
Figure 4-d Data exporting in WebHarvy----------------------------------------------------------- - 19 -
Figure 4-e Interested data to be examined-------------------------------------------------------- - 19 -
Figure 4-f Raw data (partial) ------------------------------------------------------------------------ - 20 -
Figure 4-g Using Rita WordNet to expand adjectival word lists ------------------------------ - 21 -
Figure 4-h Adjectival word lists expansion progress-------------------------------------------- - 21 -
Figure 4-i Final adjectival word list ---------------------------------------------------------------- - 21 -
Figure 4-j Hotel feature list -------------------------------------------------------------------------- - 22 -
Figure 4-k Determining the sentiment of each review------------------------------------------ - 23 -
Figure 4-l Updated data with positive or negative words and counters, as well as sentiment
result (partial)------------------------------------------------------------------------------------------ - 23 -
Figure 4-m Refined data (partial)------------------------------------------------------------------- - 24 -
Figure 5-a The Performance of Various Parisian Hotel Features ----------------------------- - 25 -
Figure 5-b The Interest of Parisian Hotel Facilities among Different Continents ---------- - 27 -
Figure 5-c The Interest of Parisian Hotel Location among Different Continents----------- - 28 -
Figure 5-d The Interest of Parisian Hotel Room among Different Continents-------------- - 29 -
Figure 5-e The Interest of Parisian Hotel Security among Different Continents ----------- - 30 -
Figure 5-f Interest of Parisian Hotel Service among Different Continents ------------------ - 31 -
Figure 5-g The Interest of Parisian Hotel Value among Different Continents -------------- - 32 -
Figure 5-h The Interest of Parisian Hotel Facilities among Different Age Groups --------- - 33 -
Figure 5-i The Interest of Parisian Hotel Location among Different Age Groups ---------- - 34 -
Figure 5-j The Interest of Parisian Hotel Room among Different Age Groups-------------- - 35 -
Figure 5-k The Interest of Parisian Hotel Security among Different Age Groups ---------- - 36 -
Figure 5-l The Interest of Parisian Hotel Service among Different Age Groups ------------ - 37 -
Figure 5-m The Interest of Parisian Hotel Value among Different Age Groups------------- - 38 -
Figure 5-n The Interest of Parisian Hotel Facilities among Different Genders ------------- - 39 -
Figure 5-o The Interest of Parisian Hotel Location among Different Genders-------------- - 40 -
Figure 5-p The Interest of Parisian Hotel Room among Different Genders ----------------- - 41 -
Figure 5-q The Interest of Parisian Hotel Security among Different Genders -------------- - 42 -
Figure 5-r The Interest of Parisian Hotel Service among Different Genders---------------- - 43 -
Figure 5-s The Interest of Parisian Hotel Value among Different Genders------------------ - 44 -
Figure 5-t Travel Frequency to Paris among Different Continents --------------------------- - 45 -
Figure 5-u Travel Frequency to Paris among Different Age Groups-------------------------- - 47 -
9. (c)C
opyrighted
- 8 -
Figure 5-v Travel Frequency to Paris among Different Genders------------------------------ - 49 -
Figure 5-w Travel Seasons to Paris----------------------------------------------------------------- - 50 -
Figure 5-x Participation in Online Review Writing between Male and Female ------------ - 51 -
Figure 5-y Participation in Online Review Writing among Different Authors-------------- - 52 -
Figure 5-z The Initial Impulse of Online Review Writing--------------------------------------- - 53 -
Figure 5-aa Positive Online Review Writing Between Male and Female -------------------- - 54 -
Figure 5-bb Negative Online Review Writing Between Male and Female------------------- - 55 -
10. (c)C
opyrighted
- 9 -
1. Introduction
Nowadays, people can travel around the world more easily than ever due to the vast
improvement in accessibility. Despite the protracted economic difficulties, Europe
has reached 534 million tourist arrivals in 2012, which is 18 million more than in
2011 and accounting for 52% of all international arrivals worldwide (UNWTO,
2013). As indicated in Figure 1a, France is the top one country in terms of
international tourist arrivals with 83 million visitors in 2012. In addition, the capital
city of France, Paris is the 3rd top visited destination city as shown in Figure 1-b.
Therefore, Paris is chosen as the designated city to be investigated in this study.
Figure 1-a International tourist arrivals (UNWTO, 2013)
Figure 1-b Global top 20 destination cities (Master Card Global Destination Cities Index, 2013)
11. (c)C
opyrighted
- 10 -
Paris, which is also known as the City of Light, is home to the famous and luxurious
fashion brands, such as Chanel, Lancôme, L'Oréal, etc. Paris has 2.2 million
inhabitants and the official language is French. Paris attracts millions of tourists
every year with its abundant iconic landmarks, such as Notre-Dame de Paris, La
Tour Eiffel, Arc de Triomphe, Musée du Louvre, Sacré-Cœur Basilica, etc (Wikitravel,
2013; Paris Digest, 2013). Thus, making it one of the top most visited cities in the
world.
Figure 1-c Iconic landmarks of Paris
A majority of the travelers would need to book a hotel when they travel to a certain
city. With the advancement of Internet, more travelers are switching to Internet for
travel planning (Litvin et al., 2008; Sigala et al., 2001). However, online hotel
booking is seen as high-risk purchase as it cannot be evaluated before consumption
(Lewis & Chambers, 2000). This is where online hotel reviews come in handy.
12. (c)C
opyrighted
- 11 -
2. Problem Statement
2.1 Insights from consumers
With the ubiquity of Internet, there is a significant growth of user generated
content posted on the websites. This information is valuable to an organization
as it shows the preferences and insights of customers. If being used strategically,
it may bring competitive advantage to the organization. The management may
use the information to better identify the needs of their customers and improve
their products or services to better accommodate the customers (Loureiro &
Kastenholz, 2011; Jun, et. al., 2010). It serves as a major source of business
intelligence (Chung & Tseng, 2012). In addition, a good knowledge of customers’
preferences and behavior can assist managers in decision making, which is a key
to business success (Rong et. al., 2012)
2.2 Phenomenon of electronic word of mouth (eWOM)
Human behaviors are changing due to the phenomenon of eWOM. The major
benefit of online reviews is they view things from a user’s perspective, thus
offering more consumer-oriented information. That is the reason why online
reviews are able to influence the purchase decisions of potential consumers.
Furthermore, online reviews break the geographical boundaries as they can be
reached far beyond the local community through the Internet. Online reviews
are measurable and easy to observe (Lee et al., 2007).
2.3 Nature of products/services
There are two types of products, tangible products such as camera or intangible
products such as services. Staying in a hotel is categorized as a service because
one has to experience it. Due to its nature, it often involves higher risks and
uncertainties. Online hotel reviews written by other consumers who have
experienced the services are seen as trusted source of information for potential
consumers. They believe it would help in reducing their uncertainties and
making a better purchase decision (Kiang et al., 2011).
13. (c)C
opyrighted
- 12 -
3. Literature Review
3.1 Online Travel Reviews
According to an industrial survey conducted by Channel Advisor (2011), 90% of
consumers read online reviews, with 83% consider their purchase behaviors are
affected by these reviews. This result is further supported by Ipsos Global where
they argued that 78% of consumers are influence by online reviews during their
purchase decision making process (eMarketer, 2013). Inarguably, consumers’
opinions can be shared and accessed easily through the Internet (Dellarocas,
2003).
Anderson (2012) pointed out the increasing numbers of travelers in consulting
online travel reviews before purchasing. Many travelers are using virtual
communities such as Trip Advisor, Virtual Tourist and Lonely Planet to gather or
provide information, compare and evaluate alternatives. This is supported by a
study done by Schindler and Bickart (2005) where they found out online
reviews are often used to gather information and to ensure a previously made
decision is correct. This is mainly due to the perceptions of traveler about online
reviews written by other travelers are more current and trustworthy (Gretzel &
Yoo, 2008).
Pan et al. (2007) believes that online travel reviews have become major sources
of information for travelers. Dickinger and Mazanec (2008) argued that the
online reviews can influence online hotel bookings. The online reviews can aid
travelers in better understanding a hotel.
14. (c)C
opyrighted
- 13 -
3.2 Research context: Hostelworld.com
Ray Nolan and Tom Kennedy founded Web Reservations International (WRI) in
1999 and created www.hostelworld.com for hostel bookings (Golden &
Cunningham, 2005). The company started off with focusing in hostel online
bookings, and now expanded its business to hotels, campsites, bed and breakfast.
Hostelworld.com lists over 27,000 properties in more than 18- countries and has
successful relationships with over 3,500 distributions partners, including
world’s leading brands such as Lonely Planet and Ryanair (hostelworld.com,
2013).
Figure 3-a Homepage of Hostelworld.com (Hostelworld.com, 2013)
3.2.1 Review at Hostelworld.com
Figure 3-b shows reviews at Hostelworld.com. There are numerical rating,
textual review, posting date, author’s identity and contribution.
16. (c)C
opyrighted
- 15 -
3.3 Natural Language Processing
3.3.1 Stop words removal
Several studies have proven the improvement of text retrieval, classification
and summarization through pre-processing steps such as stop words
removal and word stemming (Salton et al,. 1997; Yang & Chute, 1994).
According to Van Rijsbergen (1979), stop words removal aids in reducing
noisy information and improving the accuracy. Stop words are those words
which have no significant effect and meaning in a sentence, including articles,
prepositions, conjunctions and some other high-frequency words. For
example, ‘a’, ‘and’, ‘you’, ‘are’, etc. Word stemming is a technique to change
derived words back to their root forms. For instance, ‘completely’,
‘completed’, completing -> complete.
We are using the stop words list generated by the Information Retrieval
Group of University of Glasgow (The Information Retrieval Group, 2013).
Based on the study done by Liu et al. (2013), we only focus on adjectival
words as people tend to use adjectives in expressing their sentiment. As we
are focusing on adjectival words, which are a type of derived words, we will
not be using word stemming.
3.3.2 Adjectival word list expansion by WordNet
Based on the adjectival hotel word lists and hotel features generated by Xia
and Peng (2009), we further expand the adjectival word lists by using
WordNet and adjust the hotel features list to better suit our study. WordNet,
an English lexical database in which words are grouped into sets of
synonyms is implemented in this study to expand the positive and negative
word lists (WordNet, 2014). A word that is a synonym of a positive adjectival
word is added to the original positive word list, and vice versa. WordNet is
chosen in this study due to its successful implementation in previous studies
(Hu & Liu, 2004; Zhuang et al., 2006).
18. (c)C
opyrighted
- 17 -
4.2 Project Flow
Figure 4-b Project flow
Internet
Raw data
Data Pre-
processing
Sentiment
Analysis
Updated data
Refined data
19. (c)C
opyrighted
- 18 -
4.3 Data Collection
An intelligent web scraper, WebHarvy (https://www.webharvy.com/) is used to
collect online review data from Hostelworld.com as it provides user friendly
interface and offers a variety of file formats for data extracted. It is just simple
point and click. Figure 5-c shows a screen shot of WebHarvy. User just needs to
input the website address on the URL bar and click on start Config. Then, user
may point (yellow highlight will appear) and click on the desired data to be
collected. A preview of data collected is shown on the bottom part of the
application. Once Config is set up, user may click on Start Mine to run the process
automatically. When data scraping process is finished, user may export the
scraped data to either database or different file formats such as csv as depicted
in Figure 5-d.
Figure 4-c Screen shot of WebHarvy
20. (c)C
opyrighted
- 19 -
Figure 4-d Data exporting in WebHarvy
Figure 5-e shows the data we are interested in examining, that are numerical rating,
date of posting, identity of author such as name, country, gender and age,
contribution of the author in posting review, and textual review. These data are
corresponding to those shown in Figure 5.1d.
Figure 4-e Interested data to be examined
The raw data collected from WebHarvy are in csv format. These data are then
processed and save as xls format manually. Figure 5-f shows a screen shot of the
data collected that will be used for this study.
21. (c)C
opyrighted
- 20 -
Figure 4-f Raw data (partial)
4.3.1 Sampling Data
After removing redundant data, we have a total of 4500 online hotel reviews
during the time period September 2009 – January 2014 to conduct
quantitative analysis. Each review contains rating of the hotel given by the
author, profile of the author, date of posting and textual comment. The
profile of the author includes name, age, gender, country and contribution.
There are six groups of gender, namely Male, Female, All Male Group, All
Female Group, Mixed Group and Couple and four segments of age, namely
18-24, 25-30, 31-40 and 41+. The contribution displays how many comments
an author has written so far in www.hostelword.com.
22. (c)C
opyrighted
- 21 -
4.4 Data Pre-Processing
Based on a list of 318 stop words, we pre-processed our data by removing words
that bear little or no semantic meanings. Then, we used the two sets of adjectival
word lists, the positive and negative word lists generated by Xia and Peng
(2009). These lists are further extended by implementing Rita WordNet, which is
a Java API developed by Daniel Howe. We ran the program once with the initial
adjectival word lists and again but with the updated adjectival word lists. This
step is trying to minimize the word redundancies and get more accurate words.
When it is completed, we reviewed the adjectival word lists to remove
redundant words and words that bear little meaning to our focus of study.
Finally, we came up with a list of 196 words for each adjectival word lists.
Figure 4-g Using Rita WordNet to expand adjectival word lists
Figure 4-h Adjectival word lists expansion progress
Figure 4-i Final adjectival word list
23. (c)C
opyrighted
- 22 -
4.5 Data Processing
4.5.1 Hotel feature list generation
According to Lockyer (2005), a hotel’s location, price, facilities and
cleanliness have powerful impact on travelers in choosing a hotel. Travelers
are also interested in the hotel facilities, room size, breakfast and location
(Strungam et. al, 2010). On top of that, Choi and Chu (2001) revealed some of
the hotel attributes that influence travelers in selecting a hotel, which are
room quality, service quality and value. We referred to these studies and
tuned accordingly to generate a hotel feature list that best fit our study.
Figure 4-j Hotel feature list
4.5.2 Hotel feature matching
First, we split the complete review into several sentences according to dot, ‘,’.
Then we sought and compared if there is a word in the split sentence that
match our hotel feature list. If so, we continue with finding a maximum of 3
words before and after the found word to do sentiment analysis.
4.5.3 Hotel feature and review sentence sentiment analysis
We judged the sentiment by counting and comparing the number of positive
and negative words in each sentence. If the number of positive words is
greater than negative words, then the sentence is categorized as a positive
review, and vice versa. However, if the number of positive words is equal to
24. (c)C
opyrighted
- 23 -
negative words, then the sentence is categorized as a neutral review. The
sentiments of the hotel feature words and review sentences, positive or
negative words, as well as the number of positive or negative words are then
written into new columns in the excel file.
Figure 4-k Determining the sentiment of each review
Figure 4-l Updated data with positive or negative words and counters, as well as sentiment result
(partial)
25. (c)C
opyrighted
- 24 -
4.6 Data Post-Processing
We need to refine our updated data in order to smoothing the data analysis
process later. First, we changed the ‘Date’ to a recognizable excel date format
and removed the text in ‘Contribution’ so that it is recognized as number. Then
we added two new columns ‘Continents’ and ‘Season’ to categorize the countries
and dates.
Figure 4-m Refined data (partial)
26. (c)C
opyrighted
- 25 -
5 Data Analysis
5.1 The Performance of Various Parisian Hotel Features
Figure 5-a The Performance of Various Parisian Hotel Features
Observation:
Figure 5-a provides an overview of the performance of various Parisian Hotel
Features, including Facilities, Location, Room, Security, Service and Value. Overall,
travelers think that the Parisian hotels are worth the price, offer good services and
located in the heart of centre which improves their accessibilities to different places.
However, travelers believe there are rooms for improvements in the facilities and
room offered. Interestingly, travelers barely mention about their safety concerns.
Inference:
- Travelers are willing to spend their pennies provided that the hotel’s offer is
proportional to its price.
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
Facilities Location Room Security Service Value
The Performance of Various Parisian
Hotel Features
Positive
Negative
27. (c)C
opyrighted
- 26 -
- Parisian hotels offer exceptionally good services. Staffs are well trained and
they understand the importance of “customer is king”.
- It is easy to get around Paris as the hotels are strategically located close to
the attractions as well as metro stations.
- Parisian hotels need to make some improvements in their facilities and room
such as offering stable internet connection, good breakfast, clean bathroom
and room, comfortable pillows.
- Paris is a safe city to travel to where criminal rates are low.
28. (c)C
opyrighted
- 27 -
5.2 The Interest of Various Hotel Features among Different Continents
5.2.1 Hotel Facilities
Figure 5-b The Interest of Parisian Hotel Facilities among Different Continents
Observation:
Figure 5-b illustrates that travelers from North America, Oceania and South America
are highly concerned about the hotel facilities, each with over 40% of online reviews
mentioning about hotel facilities. This is followed by travelers from Europe and Asia
at 39.1% and 38.6%. In contrast, travelers from Africa show the least concerned in
hotel facilities, with only 36.8%.
35.0
36.0
37.0
38.0
39.0
40.0
41.0
42.0
Africa Asia Europe North
America
Oceania South
America
The Interest of Parisian Hotel Facilities
Among Different Continents
Facilties
29. (c)C
opyrighted
- 28 -
5.2.2 Hotel Location
Figure 5-c The Interest of Parisian Hotel Location among Different Continents
Observation:
Figure 5-c depicts that travelers from North America, Oceania, South America and
Europe shows great interest in hotel location, each with over 50% of online reviews
mentioning about hotel location whereas travelers from Asia shows little interest in
hotel location, stated at 47.8%. Similar to previous observation, travelers from
Africa show the least interest in hotel location as compared to other continents, with
only 42.1%.
35.0
37.0
39.0
41.0
43.0
45.0
47.0
49.0
51.0
53.0
55.0
Africa Asia Europe North
America
Oceania South
America
The Interest of Parisian Hotel Location
Among Different Continents
Location
30. (c)C
opyrighted
- 29 -
5.2.3 Hotel Room
Figure 5-d The Interest of Parisian Hotel Room among Different Continents
Observation:
Figure 5-d depicts that European travelers are most interested in the hotel room,
accounting to 51.2% of online reviews with regard to hotel room whereas African
travelers are least interested, with a mere 34.2%. Travelers from Oceania, North
America, South America and Asia are generally interested in the hotel room with a
number of online reviews ranging between 42.4% and 48.4%.
Inference:
- European travelers are highly concerned about hotel room when choosing a
hotel. This finding is consistent with a study done by Li et. al. (2013) where
they found out European travelers value room quality.
30.0
35.0
40.0
45.0
50.0
55.0
Africa Asia Europe North
America
Oceania South
America
The Interest of Parisian Hotel Room
Among Different Continents
Room
31. (c)C
opyrighted
- 30 -
5.2.4 Hotel Security
Figure 5-e The Interest of Parisian Hotel Security among Different Continents
Observation:
Figure 5-e shows that travelers have little interest about hotel security in general.
Despite that, North American travelers are considered as having the highest interest
of hotel security, recorded at 1.7% whereas African travelers have 0% interest
about hotel security. Travelers from Asia show an interest of 0.8%, following by
Oceania at 0.7%, Europe at 0.6% and South America at 0.5%.
Inference:
- Hotel security is believed to have little emphasis from travelers in general.
This inference is in line with a study that depicted security as bearing little
importance for travelers (Choi & Chu, 2001).
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
Africa Asia Europe North
America
Oceania South
America
The Interest of Parisian Hotel Security
Among Different Continents
Security
32. (c)C
opyrighted
- 31 -
5.2.5 Hotel Service
Figure 5-f Interest of Parisian Hotel Service among Different Continents
Observation:
Figure 5-f demonstrates the African travelers show the most interest in hotel service,
reported at a high 47.4% whereas Asian travelers show the least interest, with only
32.8%. Travelers from the other continents show general interest in hotel service,
with online reviews ranging between 40.3% and 43.8%.
Inference:
- Asian travelers show the least interest in hotel service. This finding
contradicts with a study done by Li et. al. (2013) where they revealed Asian
travelers are highly concerned about the hotel service.
30.0
32.0
34.0
36.0
38.0
40.0
42.0
44.0
46.0
48.0
50.0
Africa Asia Europe North
America
Oceania South
America
The Interest of Parisian Hotel Service
Among Different Continents
Service
33. (c)C
opyrighted
- 32 -
5.2.6 Hotel Value
Figure 5-g The Interest of Parisian Hotel Value among Different Continents
Observation:
Figure 5-g demonstrates a similar phenomenon as in Figure 5-f where the African
travelers show the most interest in hotel value, reported at an incredibly high 71.1%
whereas Asian travelers show the least interest, with a mere 48.8%. Travelers from
the other continents show common interest in hotel value, with Oceania at 61%,
Europe at 60.5%, North America at 59.7% and South America at 58.4%.
Inference:
- The group that seconded the interest for hotel value belongs to travelers
from Oceania. This result is relatively compatible a study that mentioned the
preferred criterion for Oceania travelers is value (Li et. al., 2013).
35.0
40.0
45.0
50.0
55.0
60.0
65.0
70.0
75.0
Africa Asia Europe North
America
Oceania South
America
The Interest of Parisian Hotel Value
Among Different Continents
Value
34. (c)C
opyrighted
- 33 -
5.3 The Interest of Various Hotel Features among Different Age Groups
5.3.1 Hotel Facilities
Figure 5-h The Interest of Parisian Hotel Facilities among Different Age Groups
Observation:
Figure 5-h depicts that young travelers between the ages 18-24 have the highest
concern about hotel facilities, stated at 42.4%. However, this concern about hotel
facilities gradually decreases among travelers between the ages 25-30, 31-40 and
41+, accounted for 39%, 37.8% and 35%.
30.0
32.0
34.0
36.0
38.0
40.0
42.0
44.0
18-24 25-30 31-40 41+
The Interest of Parisian Hotel Facilities
Among Different Age Groups
Facilties
35. (c)C
opyrighted
- 34 -
5.3.2 Hotel Location
Figure 5-i The Interest of Parisian Hotel Location among Different Age Groups
Observation:
Figure 5-i illustrates a similar curve as in Figure 5-h, that is young travelers between
the ages 18-24 have the highest concern about hotel location, reported at 54.6%.
However, this concern of hotel location progressively decreases among travelers
between the ages 25-30, 31-40 and 41+, accounted for 51.1%, 46.6% and 45.8%.
30.0
35.0
40.0
45.0
50.0
55.0
60.0
18-24 25-30 31-40 41+
The Interest of Parisian Hotel Location
Among Different Age Groups
Location
36. (c)C
opyrighted
- 35 -
5.3.3 Hotel Room
Figure 5-j The Interest of Parisian Hotel Room among Different Age Groups
Observation:
Figure 5-j shows that young travelers aged between 18 and 30 are more interested
in hotel room as compared to senior travelers who are aged above 30. The interest
of young travelers is around 49% and above whereas the interest of senior travelers
is around 42% and below.
38.0
40.0
42.0
44.0
46.0
48.0
50.0
18-24 25-30 31-40 41+
The Interest of Parisian Hotel Room
Among Different Age Groups
Room
37. (c)C
opyrighted
- 36 -
5.3.4 Hotel Security
Figure 5-k The Interest of Parisian Hotel Security among Different Age Groups
Observation:
Figure 5-k demonstrates that travelers commonly have very little interest in the
hotel security. Despite that, travelers between the ages 18-24 are still considered as
having the highest interest in hotel security, reported at 1.1%. The interest of hotel
security among travelers between the ages 25-30, 31-40 and 40+ accounted for
0.9%, 0.7% and 0.8%.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
18-24 25-30 31-40 41+
The Interest of Parisian Hotel Security
Among Different Age Groups
Security
38. (c)C
opyrighted
- 37 -
5.3.5 Hotel Service
Figure 5-l The Interest of Parisian Hotel Service among Different Age Groups
Observation:
Figure 5-l illustrates those travelers of age groups 18-24, 25-30 and 41+ are highly
concerned about the hotel service, each accounted for more than 40% of online
reviews about hotel service. In contrast, travelers of age group 31-40 are not as
concerned as the other groups, with only 32.5%.
30.0
32.0
34.0
36.0
38.0
40.0
42.0
44.0
18-24 25-30 31-40 41+
The Interest of Parisian Hotel Service
Among Different Age Groups
Service
39. (c)C
opyrighted
- 38 -
5.3.6 Hotel Value
Figure 5-m The Interest of Parisian Hotel Value among Different Age Groups
Observation:
Figure 5-m demonstrates that young travelers of age group 18-24 are very
concerned about the hotel value, reported at a high 63%, following by travelers of
age group 25-30 at 58.6%. However, travelers of age groups 31-40 and 41+ show
little concern about the hotel value, with only 50.2% and 49.6%.
35.0
40.0
45.0
50.0
55.0
60.0
65.0
18-24 25-30 31-40 41+
The Interest of Parisian Hotel Value
Among Different Age Groups
Value
40. (c)C
opyrighted
- 39 -
5.4 The Interest of Various Hotel Features among Different Genders
5.4.1 Hotel Facilities
Figure 5-n The Interest of Parisian Hotel Facilities among Different Genders
Observation:
Figure 5-n shows that female is highly concerned about the hotel facilities which
accounted for 44.1% whereas male is extremely unconcern about the hotel facilities,
with a mere 33.6%. Couple and mixed group are considerably concerned about the
hotel facilities, each accounted for 41.6% and 41.2%.
30.0
32.0
34.0
36.0
38.0
40.0
42.0
44.0
46.0
Couple Mixed Group Male Female
The Interest of Parisian Hotel Facilities
Among Different Genders
Facilties
41. (c)C
opyrighted
- 40 -
5.4.2 Hotel Location
Figure 5-o The Interest of Parisian Hotel Location among Different Genders
Observation:
Figure 5-o projects that couple and female have the most interest in hotel location,
each accounted for more than 50%. In contrast, mixed group and male have less
interest in hotel location, each accounted for less than 50%.
30.0
35.0
40.0
45.0
50.0
55.0
60.0
Couple Mixed Group Male Female
The Interest of Parisian Hotel Location
Among Different Genders
Location
42. (c)C
opyrighted
- 41 -
5.4.3 Hotel Room
Figure 5-p The Interest of Parisian Hotel Room among Different Genders
Observation:
Figure 5-p depicts that couple has huge interest in hotel room, which is reported as
high as 56.7%. This is followed by mixed group at 54.2% and female at 47.7%. On
the other hand, male has very little interest in hotel room, which accounted for only
40.9%.
Inference:
- Couples have a significant amount of care regarding hotel room. This result
matches the findings of Li et. al. (2013).
30.0
35.0
40.0
45.0
50.0
55.0
60.0
Couple Mixed Group Male Female
The Interest of Parisian Hotel Room
Among Different Genders
Room
43. (c)C
opyrighted
- 42 -
5.4.4 Hotel Security
Figure 5-q The Interest of Parisian Hotel Security among Different Genders
Observation:
Figure 5-q demonstrates that travelers generally are not very concern about the
hotel security. But at 2.3%, mixed group is portrayed as greatly aware of safety
concern in comparison to the other groups which are around or lower than 1%.
0.0
0.5
1.0
1.5
2.0
2.5
Couple Mixed Group Male Female
The Interest of Parisian Hotel Security
Among Different Genders
Security
44. (c)C
opyrighted
- 43 -
5.4.5 Hotel Service
Figure 5-r The Interest of Parisian Hotel Service among Different Genders
Observation:
Figure 5-r shows that female, couple and mixed group are generally concerned
about the hotel service as each of them has more than 40% of online reviews with
hotel service mentioned while male has only 36.7% of online reviews with hotel
service mentioned.
30.0
32.0
34.0
36.0
38.0
40.0
42.0
44.0
46.0
Couple Mixed Group Male Female
The Interest of Parisian Hotel Service
Among Different Genders
Service
45. (c)C
opyrighted
- 44 -
5.4.6 Hotel Value
Figure 5-s The Interest of Parisian Hotel Value among Different Genders
Observation:
Figure 5-s shows that mixed group has an exceptionally high interest in hotel value,
stated at 62.5%. This is followed by female at 59.5%, couple at 57.6% and male at
57.1%.
30.0
35.0
40.0
45.0
50.0
55.0
60.0
65.0
Couple Mixed Group Male Female
The Interest of Parisian Hotel Value
Among Different Genders
Value
46. (c)C
opyrighted
- 45 -
5.5 Travel Frequency to Paris
5.5.1 Among Different Continents
Figure 5-t Travel Frequency to Paris among Different Continents
Observation:
Figure 5-t illustrates that the European traveler is the largest group to travel to
Paris, which is as high as 37.1% whereas African traveler is the smallest group to
travel to Paris, which is as low as 0.8%. Following European, North American
traveler is the second largest group to travel to Paris, at 26.5%. This is followed by
Asian traveler at 14.4%, Oceania traveler at 12.7% and South American traveler at
8.4%.
Inference:
- With the advancement of railway system in Europe, it is convenient for
Europeans to travel within Europe. For instance, a comprehensive and
modern rail network offer by the French, Italian and Spanish railway
0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0%
Africa
South America
Oceania
Asia
North America
Europe
Travel Frequency to Paris Among
Different Continents
Total
47. (c)C
opyrighted
- 46 -
companies have enhanced the mobility of European within Europe (Eurail,
2014). This explains why European traveler is the largest group to travel to
Paris.
48. (c)C
opyrighted
- 47 -
5.5.2 Among Different Age Groups
Figure 5-u Travel Frequency to Paris among Different Age Groups
Observation:
Figure 5-u illustrates that the young travelers of age group 18-24 which accounts for
42.7% traveled the most to Paris. The age group 25-30 that accounts for 35.5% is
not far behind. However, there is a huge gap between young traveler groups and
senior traveler groups. Travelers that traveled to Paris of age groups 31-40 and 40+
are as low as 13.1% and 8.6%.
Inference:
- Young people aged below 30 are eager to explore the world and have the
energy to travel around. Furthermore, they have the freedom to move
around as most of them have not start up a family. Students exchange
programme and working holiday have also encouraged youngsters to travel
more often.
0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0% 45.0%
41+
31-40
25-30
18-24
Travel Frequency to Paris Among
Different Age Groups
Total
49. (c)C
opyrighted
- 48 -
- Senior people aged 31 and above are likely to be in their career or started up
a family. Thus they have limited time to go on travel. Even though they have
time and money, most of them prefer staying home than traveling around as
they think traveling would be too exhausting.
50. (c)C
opyrighted
- 49 -
5.5.3 Among Different Genders
Figure 5-v Travel Frequency to Paris among Different Genders
Observation:
As many as 41% of females traveled to Paris, this makes them the largest group to
travel to Paris as shown in Figure 5-v. This is followed by male group at 32.7% and
couple group at 19.6%. By contrast, mixed group is the smallest group to travel to
Paris, with only 6.7% in total.
Inference:
- Females are traveling more often than ever. This is in line with previous
study stating the increase engagement of women in business travel
(Brownell, 2011). On the other hand, Paris is the home to fashions. This could
be the reason that a large amount of female is traveling to Paris.
41.0%
32.7%
19.6%
6.7%
Travel Frequency to Paris Among
Dfifferent Genders
Female
Male
Couple
Mixed Group
51. (c)C
opyrighted
- 50 -
5.6 Travel Seasons to Paris
Figure 5-w Travel Seasons to Paris
Observation:
Figure 5-w projects summer as the top season with the most travelers in Paris,
which accounted for 35.5% whereas winter is the lowest season with the least
travelers, reported at only 16.9%. Fall and spring are not far behind from summer,
with 27.7% and 19.9%.
Inference:
- Summer is the best season to travel to Paris as there are many festivals and
concerts happening during the summer and the weather is warm (AOL
Travel, 2014).
- Travelers mostly travel during the summer as most of the students are
having long summer holiday.
0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0%
Winter
Spring
Fall
Summer
Travel Seasons to Paris
Total
52. (c)C
opyrighted
- 51 -
5.7 Online Review Writing
5.7.1 Participation in Online Review Writing Between Male and Female
Figure 5-x Participation in Online Review Writing between Male and Female
Observation:
Figure 5-x illustrates the online reviews community is dominated by females,
reported at a high 55.7% whereas males comprised of only 44.3%.
Inference:
- The online reviews community is prevailed by females. This outcome is
supported by several recent studies that show women predominate the
Internet (Kim et. al, 2007; Toh et. al., 2011).
44.3%
55.7%
Participation in Online Review Writing
Between Male and Female
Male
Female
53. (c)C
opyrighted
- 52 -
5.7.2 Participation in Online Review Writing Among Different Authors
Figure 5-y Participation in Online Review Writing among Different Authors
Observation:
Figure 5-y depicts that the majority of travelers remained anonymous in writing
online reviews, which is as high as 84%. In contrast, only 16% of travelers
remained non-anonymous.
Inference:
- Travelers prefer to remain anonymous when writing online reviews. They
may not want their privacy being invaded and also to avoid judgment by the
others (Goodwin, 1992).
- Women are believed to possess greater concern about privacy and less likely
to disclose identity information (Fogel & Nehmad, 2009). As our study
showed that there are more females than males in online review writing,
thus there is a significant amount of anonymous authors.
16.0%
84.0%
Participation in Online Review Writing
Among Different Authors
Non Anonymous
Anonymous
54. (c)C
opyrighted
- 53 -
5.7.3 The Initial Impulse of Online Review Writing
Figure 5-z The Initial Impulse of Online Review Writing
Observation:
Figure 5-z projects the number of positive online reviews is extremely high for
author who wrote only one review, reported at 85.2% whereas the number of
negative online reviews is unremarkably low, with only 14.8%.
Inference:
- The initial impulse of writing an online review is more likely to be triggered
by positive experience. This implies there is a significant amount of positive
online reviews than negative online reviews. This implication is supported by
a recent study that concluded online hotel reviews is dominated by positive
online reviews (Melian et. al., 2013).
14.8%
85.2%
The Initial Impulse of Online Review
Writing
Negative
Positive
55. (c)C
opyrighted
- 54 -
5.7.4 Positive Online Review Writing Between Male and Female
Figure 5-aa Positive Online Review Writing Between Male and Female
Observation:
Figure 5-aa illustrates the number of positive online reviews writing between male
and female is comparable. Female is reported at 50.7% whereas male is reported at
49.3%.
Inference:
- Both male and female like to share their positive experiences with others,
hoping that the others would also enjoy the positive experiences like them. In
addition, when they are satisfied with the hotels, they would help to promote
the hotels by recommending to the others (Sundaram et. al., 1998).
49.350.7
Positive Online Review Writing Between
Male and Female
Male
Female
56. (c)C
opyrighted
- 55 -
5.7.5 Negative Online Review Writing Between Male and Female
Figure 5-bb Negative Online Review Writing Between Male and Female
Observation:
Male is reported writing more negative online review than its female counterparts
as shown in Figure 5-bb, which accounted for 54.6%. In contrast, only 45.4% of
female wrote negative online reviews.
Inference:
- Males post more negative reviews. The reason may be to express their
negative emotions and exercise their consumer rights in addition to warning
others (Bronner & Hoog, 2011).
54.6
45.4
Negative Online Review Writing Between
Male and Female
Male
Female
57. (c)C
opyrighted
- 56 -
6 Limitations
6.1 Limited domain and sentiment corpus
The research of sentiment analysis within the hotel industry is pretty
limited. Therefore, the sentiment corpus may not be content specific and not
that accurate. For example, ‘helpful’, a positive word makes sense for
describing ‘staff’, but may not be the same case for describing ‘location’.
6.2 Limited data set
Due to the limited amount of time, we focused our study on Parisian hotels.
Many reviews extracted from www.hostelworld.com which are incomplete
and not written in English are removed, thus reducing the number of data
available for analysis.
6.3 Unilingual
This study only focused on English written reviews. We may miss out some
important data which are written in other languages such as Chinese,
Spanish, German, etc.
58. (c)C
opyrighted
- 57 -
7 Future Work
7.1 Develop more complicated sentiment algorithm
We may consider improving our sentiment algorithm to include the degree of
sentiments in the future. For instance, words like ‘very’, ‘somewhat’, ‘a little’,
‘many’, etc. may influence the sentiment result. In addition to that, we should
take negations into consideration, such as the word ‘not’ which may change a
positive sentiment to negative.
7.2 Expand domain of study
We may expand our research to different cities in France and around the
world such as Bangkok, New York, etc. to make comparisons and get a global
outlook. It would be interesting to see the results of such global comparisons.
7.3 Improve method of data analysis
It can be seen that most of our results are comparable to previous studies.
However, it is hard to judge due to our simplified analysis. Therefore, we may
consider more advanced analysis method such as doing cross referencing to
find out the interrelationships among the data.
7.4 Consider Chinese review websites and data set
Our current study focused on English written reviews and used a website
which is popular among the English speaking travelers. We may consider
Chinese written reviews in the future by examining a Chinese online review
website like www.taobao.com.
59. (c)C
opyrighted
- 58 -
8 Recommendations
8.1 Improve hotel facilities and room
This study indicated that there are rooms for improvement in terms of hotel
facilities and room. Hotel facilities and room are significant for Americans,
Oceania and European travelers. Hotel managers should emphasize on
improving these aspects in order to appeal to these groups. For instance,
provide stable internet connection, clean bedding stuff, good breakfast, etc.
8.2 Focus on travelers between ages 18-30
As our result showed that the biggest group of travelers is between ages 18-
30, marketing strategies should be carefully designed with a focus on this
group of travelers. In order to mitigate the congestions during the summer
periods and to boost business during off seasons, hotel managers should
cater different package deals such as cheaper all in one package that
combines hotel stays and visiting tourist attractions.
8.3 Attract Chinese travelers
Asians are depicted as the third largest group to travel to Paris. According to
Cripps (2013), Chinese travelers are the world’s biggest spenders and they
focus on luxury shopping during their travels. Hotel managers should utilize
the characteristic of Paris as a home to famous luxury brands and
incorporate it into its marketing strategies to attract Chinese travelers.
8.4 Consider negative reviews too
Despite this study showed that there is a significant amount of positive
reviews as compared to negative reviews, hotel managers should not neglect
the influence of negative reviews. They should evaluate the negative reviews
and provide appropriate response.
60. (c)C
opyrighted
- 59 -
9 Conclusion
This study intended to illustrate the importance of electronic word of mouth and
assist hotel managers in understanding travelers’ behaviors as well as the
performance of their hotels by viewing problems from multi dimensional business
perspectives. First, the performance of Parisian hotels is analyzed. Then, the
travelers’ behaviors are compared among continents, age groups and genders.
Extensive results are generated such as tourism in Paris, preferences of travelers in
choosing hotels and travelers’ behaviors in online review writing. These results are
valuable and could be used by hotel managers in decision and strategy making. In
conclusion, electronic word of mouth is an importance source of business
intelligence and should be utilized strategically in order to generate business values.
61. (c)C
opyrighted
- 60 -
10 References
World Tourism Organization Network, (2013) ‘UNWTO Tourism Highlights’, [Pdf]
Available at:
<http://dtxtq4w60xqpw.cloudfront.net/sites/all/files/pdf/unwto_highlights13_en_
hr_0.pdf> [5th September 2013].
Wong, Y.H. & Chong, D. (2013) ‘MasterCard Global Destination Cities Index’, [Pdf]
Available at: <http://newsroom.mastercard.com/wp-
content/uploads/2013/05/Updated-Mastercard_GDCI_Final_V4.pdf> [5th
September 2013].
Wikitravel: Paris (2013) [Online] Available at: <http://wikitravel.org/en/Paris> [6th
September 2013].
Paris Digest: ParisFacts (2013) [Online] Available at:
<http://www.parisdigest.com/information/facts.htm> [6th September 2013].
Kiang, M.Y., Ye Q., Hao, Y., Chen, M. and Li, Y. (2011) ‘A Service-oriented Analysis of
Online Product Classification Methods’, Decision Support Systems, vol. 52, pp. 28-39.
Lee, J., Park, D. & Han, I. (2007) ‘The Effect of Negative Online Consumer Reviews on
Product Attitude: An Information Processing View’, Electronic Commerce Research
and Applications, vol. 7, pp. 341-352.
Chung, W. & Tseng, T. (2012) ‘Discovering Business Intelligence From Online
Product Reviews: A Rule-induction Framework’, Expert System with Applications,
vol. 39, pp. 11870-11879.
Golden, W. & Cunningham, J. (2005) ‘Case Study: Hostelworld.com’, pp. 1-9.
Hostelworld.com (2013) [Online] Available at: <
http://www.hostelworld.com/aboutus.php> [26th December 2013].
62. (c)C
opyrighted
- 61 -
Litvin, S.W., Goldsmith, R.E. & Pan, B. (2008) ‘Electronic Word-of-Mouth in
Hospitality and Tourism Management’, Tourism Management, vol. 29, no. 3, pp.
458–468.
Sigala, M., Lockwood, A. & Jones, P. (2001) ‘Strategic Implementation and IT: Gaining
Competitive Advantage from the Hotel Reservations Process’, International Journal
of Contemporary Hospitality Management, vol. 13, no. 7, pp. 364–371.
Lewis, R.C. & Chambers, R.E. (2000) Marketing Leadership in Hospitality,
Foundations and Practices (3rd edition), New York: Wiley.
Dellarocas, C. (2003) ‘The Digitization of Word-of-Mouth: Promise and Challenges of
Online Feedback Mechanisms’, Management Science, vol. 49, no. 10, pp. 1407–1424.
Anderson, C.K. (2012) ‘The Impact of Social Media on Lodging Performance’, Cornell
Hospitality Report, vol. 12, no. 5, pp. 4–12.
Schindler, R. & Bickart, B. (2005) ‘Published ‘Word of Mouth’: referable, Consumer-
generated Information on the Internet. In: Hauvgedt, C., Machleit, K., Yalch, R. (Eds.),
Online Consumer Psychology: Understanding and Influencing Behavior in the
Virtual World. Lawrence Erlbaum Associates, pp. 35–61.
Gretzel, U. & Yoo, K. (2008) ‘Use and Impact of Online Travel Reviews’, In P.
O’Connor, W. Höpken, & U. Gretzel (Eds.), Information and Communication
Technologies in Tourism 2008 (pp. 35–46) New York: Springer-Verlag.
Pan, B., MacLaurin, T. & Crotts, J.C. (2007) ‘Travel Blogs and the Implications for
Destination Marketing’, Journal of Travel Research, vol. 46, no. 1, pp. 35–45.
Dickinger, A. & Mazanec, J. (2008) ‘Consumers’ Preferred Criteria for Hotel Online
Booking. In P. Peter O’Connor, W. Wolfram Höpken, U. Ulrike Gretzel (Eds.),
Information and Communication Technologies in Tourism (pp. 244-254) New York:
Springer-Verlag.
63. (c)C
opyrighted
- 62 -
Hotels.com (2013) [Online] Available at: <http://press.hotels.com/en-us/company-
overview/hotels-com-backgrounder/> [10th November 2013].
ChannelAdvisor. (2011) ‘2011 Consumer Survey: Global Consumer Shooping Habits’.
eMarketer. (2013) ‘Users Seek Out the Truth in Online Reviews: Consumers Look for
Indicators for Review Authenticity’.
Vermeulen, I.E. & Seegers, D. (2009) ‘Tried and Tested: The Impact of Online Hotel
Reviews on Consumer Consideration’, Tourism Management, vol. 30, no. 1, pp. 123-
127.
Salton, G., Singhal, A., Mitra, M. & Buckley, C. (1997) ‘Automatic Text Structuring and
Summarization’, Information Processing and Management, vol. 33, no. 2, pp. 193-
207.
Van Rijsbergen, C.J. (1979) ‘Information Retrieval 2nd Edition;, London, UK:
Butterworths.
Liu, Y., Jin, J., Ji, P., Harding, J.A. & Fung, R.Y.K (2013) ‘Identifying Hellpful Online
Reviews: A Product Designer’s Perspective’, Computer-Aided Design, vol. 45, pp.
180-194.
The Information Retrieval Group – Stop Words List (2013) [Online] Available at: <
http://ir.dcs.gla.ac.uk/resources/linguistic_utils/stop_words> [26th December 2013].
Xia, H. & Peng, L. (2009) ‘SVM-Based Comments Classification and Mining of Virtual
Community: For Case of Sentiment Classification of Hotel Reviews’, Proceedings of
the International Symposium on Intelligent Information Systems and Applications,
pp. 507-511.
WordNet (2014) [Online] Available at: <http://wordnet.princeton.edu/> [4th
January 2014].
64. (c)C
opyrighted
- 63 -
Hu, M. & Liu, B. (2004) ‘Mining Opinion Features in Customer Reviews’, In:
Proceedings of the 19th National Conference on Artificial Intelligence, San Jose, USA,
pp. 755-760.
Zhuang, L., Jing, F. & Zhu, X.Y. (2006) ‘Movie Review Mining and Summarization’, In:
Proceedings of the 15th ACM International Conference on Information and
Knowledge Management, Arlington, Virginia, USA: ACM New York, NY, USA, pp. 43-
50.
Lockyer, T. (2005) ‘Understanding the Dynamics of the Hotel Accommodation
Purchase Decision’, International Journal of Contemporary Hospitality Management,
vol. 17, no. 6, pp. 481-492.
Stringam, B.B., Gerdes, J. & Vanleeuwen, D.M. (2010) ‘Assessing the Importance and
Relationships of Ratings on User-generated Traveler Reviews’, Journal of Quality
Assurance in Hospitality and Tourism, vol. 11, no. 2, pp. 73-92.
Choi, T.Y. & Chu, R. (2001) ‘Determinants of Hotel Guests’ Satisfaction and Repeat
Patronage in the Hong Kong Hotel Industry’, International Journal of Hospitality
Management, vol. 20, no. 3, pp. 277-297.
Rong, J., Vu, H.Q., Law, R. & Li, G. (2012) ‘A Behavioral Analysis of Web Sharers and
Browsers in Hong Kong using Targeted Association Rule Mining’, Tourism
Management, vol. 33, no. 4, pp. 731-740.
Li, G., Law, R., Vu, H.Q. & Rong, J. (2013) ‘Discovering the Hotel Selection Preferences
of Hong Kong Inbound Travelers Using the Choquet Integral’, Tourism Management,
vol. 36, pp. 321-330.
Loureiro, S.M.C. & Kastenholz, E. (2011) ‘Corporate Reputation, Satisfaction, Design
and Loyalty towards Rural Lodging Units in Portugal’, International Journal of
Hospitality Management, vol. 30, pp. 575-583.
65. (c)C
opyrighted
- 64 -
Jun, S.H., Vogt, C.A. & MacKay, K.J. (2010) ‘Online Information Search Strategies: A
Focus on Flights and Accommodations’, Journal of Travel Tourism Marketing, vol. 27,
pp. 579-595.
Bronner, A.E. & Hoog, R. (2011) ‘Vacationers and eWOM: Who Posts, and Why,
Where and What?’, Journal of Travel Research, vol. 50, pp. 15-26.
Melian, S.G., Bulchand, J.G. & Lopez, B.G.V. (2013) ‘Online Customer Reviews of
Hotels: As Participation Increases, Better Evaluation is Obtained’, Cornell Hospitality
Quarterly, vol. 54, no.3, pp. 274-283.
Toh, R.S., DeKay, C.F. & Raven, P. (2011) ‘Travel Planning: Searching for and Booking
Hotels on the Internet’, Cornell Hospitality Quarterly, vol. 52, no. 4, pp. 388-398.
Kim, D., Lehto, Y. & Morrison, A.M. (2007) ‘Gender Differences in Online Travel
Information Search: Implications for Marketing Communications on the Internet’,
Tourism Management, vol. 28, pp. 423-433.
Brownell, J. (2011) ‘Creating Value for Women Business Travelers: Focusing on
Emotional Outcomes’, Cornell Hospitality Reports, vol. 11, no. 12.
Sundaram, D.S., Mitra, K. & Webster, C. (1998) ‘Word-of-Mouth Communications: A
Motivational Analysis’, Advances in Consumer Research, vol. 25, pp. 527-531.
Eurail Unveils Mediterranean Deals. (2014) ‘Manufacturing Close – Up’, [Online]
Available at: <http://search.proquest.com/docview/1508321982?accountid=16210>
[9th April 2014].
Goodwin, C. (1992) ‘A Conceptualization of Motives to Seek Privacy for Nondeviant
Consumption’, Journal of Consumer Psychology, vol. 1, no. 3, pp. 261-284.
Fogel, J. & Nehmad, E. (2009) ‘Internet Social Network Communities: Risk Taking,
Trust and Privacy Concerns’, Computers in Human Behavior, vol. 25, pp. 153-160.
66. (c)C
opyrighted
- 65 -
AOL Travel. (2014) ‘Paris When to Go and Weather’, [Online] Available at: <
http://travel.aol.com/travel-guide/europe/france/paris-best-time-to-visit/> [9th
April 2014].
Cripps, K. (2013) ‘Chinese Travelers the World’s Biggest Spenders’, CNN, [Online]
Available at: <http://edition.cnn.com/2013/04/05/travel/china-tourists-spend/>
[9th April 2014].
68. (c)C
opyrighted
- 67 -
s1 = s1.toLowerCase();
// Scan through each word in review text and compare to each Stop Word
for (int i = 0; i < stopWrds.length; i++) {
if (s1.equals(stopWrds[i])) {
flag = 0;
}
}
// If it is not a Stop Word, then add to textOutput
if (flag != 0) {
textOutput = textOutput + s1 + " ";
}
}
textScanner.close();
return textOutput;
}
11.2 Hotel Feature Matching
public Excelnew featureListCheck (String reviewWithoutStopwordsinput,String
filepath){
Excelnew objfeature = new Excelnew();
// Strings used to build the output
String sfPosword="";
String sfNegword ="";
String compare;
int sfPoscounter=0;
int sfNegcounter=0;
try {
// Review sentence splitting
String[] sentence = reviewWithoutStopwordsinput.split(".");
for (String splitted : sentence) {
int wordstotal = countWordsTotal(splitted);
System.out.println(splitted+" |Words: " +wordstotal);
int wordcounter = 0;
Scanner textScanner = new Scanner(splitted);
while (textScanner.hasNext()) {
BufferedReader br = new BufferedReader(new
FileReader(filepath));
String nextWordofReview = textScanner.next();
wordcounter+=countWordsTotal(nextWordofReview);
// Comparing each word of review sentence with each hotel feature
word
while ((compare = br.readLine()) != null) {
if (compare.equals(nextWordofReview)) {
String words = null;
System.out.println(" found Word: " +
nextWordofReview+" Wordcounter: " +wordcounter + "
words off total "+wordstotal );
69. (c)C
opyrighted
- 68 -
// Get the word/s (max: 3) before and after the hotel
feature word in the review sentence
words = suroundingword(splitted, nextWordofReview,
wordcounter, wordstotal);
System.out.println(" featurelist
call ......."+words+"..... compareword: " +compare);
objfeature = sentimentAnalysis(words);
if(objfeature.getPosWord()!=null){
sfPosword+=", "+compare+": "+objfeature.getPosWord();
}
if (objfeature.getNegWord()!=null){
sfNegword +=", "+compare+": "+objfeature.getNegWord();
}
sfPoscounter += (objfeature.getPosResult());
sfNegcounter += (objfeature.getNegResult());
}
}
br.close();
}
}
} catch (FileNotFoundException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
objfeature.setPosWord(sfPosword);
objfeature.setPosResult(sfPoscounter);
objfeature.setNegWord(sfNegword);
objfeature.setNegResult(sfNegcounter);
return objfeature;
}
11.3 Sentiment Analysis
public Excelnew sentimentAnalysis (String reviewSentence){
// Create an object called objPosNeg
Excelnew objPosNeg = new Excelnew();
Scanner textScanner = new Scanner(reviewSentence);
// Initialize positive and negative counters
int pos = 0;
int neg = 0;
String result = null;
String pWords = null;
String nWords = null;
String positivecompare;
String negativecompare;
try {
while (textScanner.hasNext()) {
// Read positive and negative word text files
70. (c)C
opyrighted
- 69 -
BufferedReader br = new BufferedReader(new
FileReader("C:/Users/EleanorYY/Desktop/FinalYearProject/Da
ta/PositiveWords_Complete.txt"));
BufferedReader br2 = new BufferedReader(new
FileReader("C:/Users/EleanorYY/Desktop/FinalYearProject/Da
ta/NegativeWords_Complete.txt"));
String nextWordofReview = textScanner.next();
while ((positivecompare=br.readLine())!=null){
// If positive word equals to review word and pWords
doesn't equals to null (starting from 2nd round)
// Add the positive word to pWords
// Otherwise, add the positive word to pWords (only
for 1st round)
if (positivecompare.equals(nextWordofReview)){
if (pWords!=null){
pWords = pWords +", "+ positivecompare;
}else{
pWords=positivecompare;
}
pos++;
}
}
br.close();
while ((negativecompare = br2.readLine())!=null){
// If negative word equals to review word and nWords
doesn't equals to null (starting from 2nd round)
// Add the negative word to nWords
// Otherwise, add the negative word to nWords (only
for 1st round)
if (negativecompare.equals(nextWordofReview)){
if (nWords!=null){
nWords = nWords +", "+ negativecompare;
}else{
nWords=negativecompare;
}
neg++;
}
}
br2.close();
}
// Determine result by comparing positive and negative word
counters
if (pos > neg)
result = "Positive";
if (pos < neg)
result = "Negative";
if (pos == neg)
result = "Neutral";
objPosNeg.setPosResult(pos);
objPosNeg.setNegResult(neg);
objPosNeg.setEvaluationResult(result);
