Association for Information SystemsAIS Electronic Library (AISeL)

PACIS 2017 Proceedings Pacific Asia Conference on Information Systems(PACIS)

Summer 7-19-2017

Opinion Mining from Online Reviews: ConsumerSatisfaction Analysis with B&B HotelsYetian ChenNanjing University of Aeronautics and Astronautics, bigsillybear@163.com

Lin XiaoNanjing University of Aeronautics and Astronautics, xiaolin@nuaa.edu.cn

Chuanmin MiCollege of Economics and Management, Nanjing University of Aeronautics and Astronautics Nanjing 210016, China,michuanmin@163.com

Follow this and additional works at: http://aisel.aisnet.org/pacis2017

This material is brought to you by the Pacific Asia Conference on Information Systems (PACIS) at AIS Electronic Library (AISeL). It has beenaccepted for inclusion in PACIS 2017 Proceedings by an authorized administrator of AIS Electronic Library (AISeL). For more information, pleasecontact elibrary@aisnet.org.

Recommended CitationChen, Yetian; Xiao, Lin; and Mi, Chuanmin, "Opinion Mining from Online Reviews: Consumer Satisfaction Analysis with B&BHotels" (2017). PACIS 2017 Proceedings. 81.http://aisel.aisnet.org/pacis2017/81

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Opinion Mining from Online Reviews: Consumer Satisfaction Analysis with B&B

Hotels Completed Research Paper

Yetian Chen College of Economics & Management, Nanjing University of Aeronautics and

Astronautics, Nanjing, China bigsillybear@163.com

Lin Xiao College of Economics & Management, Nanjing University of Aeronautics and

Astronautics, Nanjing, China xiaolin@nuaa.edu.cn

Chuanmin Mi

College of Economics & Management, Nanjing University of Aeronautics and Astronautics, Nanjing, China

michuanmin@163.com

Abstract

Given the enormous growth and significant impact of user generated content in online hotel reviews, this study aims to mining the determinants of consumer satisfaction with B&Bs and build a hierarchical structure of these determinants. Content analysis was conducted based on the consumer review data from two well-known hotel booking websites. Ten determinants of customer satisfaction were identified. The interpretive structural modeling (ISM) technique was then used to develop a five-level hierarchical structural model based on these determinants to illustrate the influencing paths. Finally, the cross-impact matrix multiplication applied to classification (MICMAC) technique was used to analyze the driver and dependence power for each determinant. This study has the potential to make significant contributions from both the theoretical and practical perspectives in this research area.

Keywords: B&B, consumer satisfaction, opinion mining, ISM, hierarchical structural model

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Introduction

Online booking and reviewing are becoming more popular for customers with the rapid development of information technology (Xu and Li 2016). People can also use online technology to share their travel experiences through different platforms. The tremendous growth of social media and consumer-generated content on the Internet has inspired the development of the so-called big data analytics to understand and solve real-life problems (Xiang et al. 2014). Big data analytics opens the door to numerous opportunities to develop new knowledge to reshape our understanding of the field and to support decision making in the hospitality industry (Xiang et al. 2014).

Bed-and-breakfast establishments (B&B) present a unique sector within the tourism industry. These are “types of accommodation where visitors or guests pay to stay in private homes, where interaction takes place with a host and/or family usually living on the premises and with whom public space is, to a degree, shared” (Lynch 2005, pp. 534–535). In the past 20 years, the popularity and number of B&Bs has increased dramatically throughout the world, especially in rural areas. According to statistical reports (Suzhouminsu 2016), there were 42,658 B&Bs and 1.5 million rural homestays available at the end of 2015 in Mainland China alone, an increase of 42.19% compared to the second quarter of 2014. A recent study has shown that 79% of leisure travelers intend to stay at a B&B (Turner 2011). Additional growth in the number of B&B operations is expected for the next several years, particularly in rural areas where the development of hotel and motel businesses may not be feasible (Ren et al. 2015).

The increasing number of B&Bs, the ease of booking hotels online, and the vast amount of online reviews available has made competition in the tourism industry more intense than ever before (Lin et al. 2012). Moreover, many hotels offer essentially homogeneous products and services, which drives the desire of hotels to distinguish themselves from their competitors (Xiang et al. 2014). As such, guest satisfaction has been recognized as one of most important and efficient ways to succeed in the tourism industry (Anderson and Sullivan 1993). Anderson and Sullivan (1993) have stated that a firm’s future profitability depends on satisfying current customers. Recognizing this, a plethora of studies have been conducted with the aim to understand the antecedents of guest satisfaction (e.g., Choi and Chu 2001; Zhou et al. 2014). However, the majority of these studies focus on luxury, budget, or chain hotels, rather than B&Bs. These research findings are not suitable for the B&B industry and cannot serve as a reference for B&B planning and management decision-making (Chen 2015), since B&Bs are different to budget or chain hotels.

In addition, conventional studies exploring consumer satisfaction with hotels tend to use surveys as their research methods, which are regarded as high-cost and difficult to operate (Zhou et al. 2014). Researchers are increasingly realizing that tourists’ online reviews represent a rich vein of data that can contribute to the understanding of consumer perceptions and behaviors (Kim and Hardin 2010). This consumer-generated content on the internet has inspired the development of new approaches to understanding consumer satisfaction. Although some researchers have touched on this point by analyzing online review data (e.g., Xiang et al. 2014; Xu and Li 2016) with various techniques, the factors identified in their studies are treated as isolated from each other, without explaining why and how the influence of these factors on satisfaction may occur. In other words, the influencing process or paths of various factors influencing a customer’s satisfaction remains unknown since the linkages between these factors are yet to be uncovered.

To fill these gaps, this study employed the online customer reviews of B&B properties to understand customer satisfaction toward B&Bs from a comprehensive and structural perspective. Specifically, this study has two main objectives:

1. To explore the factors influencing consumer satisfaction toward B&Bs based on user-generated reviews;

2. To identify the interrelationships between these factors and build a hierarchical structure based on these factors to indicate the influencing process.

The rest of this paper is structured as follows. The definition of B&Bs and literature related to customer satisfaction are summarized first, followed by the data collection and analysis methods, Then the results of the analysis are presented, followed by discussion. Finally the paper concludes with implications, limitations, and future directions.

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Literature Review

B&B Establishments

In the tourism sector, the commonest form of accommodation other than hotels and guesthouses is the B&B establishment, a concept that originated in Europe (Nuntsu et al. 2004). It refers to a small establishment, offering a noncommercial homelike setting and serving breakfast as the only meal (Zane 1997). Lynch (2005) extends the meaning of the term and emphasizes it as types of accommodation where visitors or guests pay to stay in private homes, where interaction takes place with the host and/or family. Now it is known as a small lodging facility offering overnight accommodation and breakfast, but not usually offering other meals (Kuo et al. 2014). B&Bs allow travelers to seek lodging for the night, especially when hotels and inns are unavailable in remote areas (e.g., near the top of a mountain or on a small island). Although there is no uniform definition of a B&B, most researchers agree that the basic criteria for distinguishing B&Bs are: 1) they are family-run; 2) they are small-scale; 3) there is a certain degree of communication with the owner; and (4) they provide special services (e.g., Chen et al. 2013; Kuo et al. 2014; Zane 1997).

Customer Satisfaction

Customer satisfaction is an evaluation based on a comparison between customers’ experiences and their initial expectations (Grönroos 1983). Enhancing customer satisfaction is widely recognized as a key strategy leading to success for companies in tourism industries (Barsky 1992; Choi and Chu 2001). Recognizing this, research into customer satisfaction in the hotel industry has increased dramatically in recent years. A number of studies have focused on exploring the factors that contribute to customer satisfaction in various hotel contexts, such as starred hotels, budget hotels, and chain hotels. Among these studies, the quantitative approach with surveys was extensively adopted by researchers to verify the multiple determinants of customer satisfaction (e.g., Choi and Chu 2001; Deng et al. 2013; Kim et al. 2006).

In recent years, a large amount of customer reviews are available on the hotel booking websites. This user generated content is informative since consumers are free to express their feelings and comment on their satisfaction after leave the hotels. Thus, researchers started to pay attention to these reviews and utilized various methods to analyze the online reviews to explore factors influencing customer satisfaction. For instance, Zhou et al. (2014) used content analysis based on 1,345 reviews on agoda.com and found that customer satisfaction was influenced by location, physical setting (food, hotel and room), staff, and value. Xu and Li (2016) utilized text mining to 3,480 reviews on booking.com for four different types of hotels and indicated that good location and view, friendly staff, nice clean rooms, and good restaurant were significantly evaluated by customers. Xiang et al. (2014) adopted a text analytics process to analyze the 60,648 customer reviews from expedia.com and discovered that core product, family friendliness, and staff were critical factors of customer satisfaction.

Although these studies have uncovered a comprehensive list of factors that can influence consumer satisfaction toward hotels (as shown in Table 1), there remain several gaps to be fulfilled. First, the majority of studies employed a quantitative research approach by using surveys and qualitative studies based on user-generated content are still limited. The vast amount of user-generated data online can provide more complete and reliable information reflecting consumer perceptions and behavior. Thus, qualitative studies exploring determinants of customer satisfaction based on user-generated content can offer greater theoretical and practical value. Second, most of studies focused on starred hotels and budget hotels, while B&Bs are ignored in the literature. As a unique style of accommodation different from other types of hotels, it is inappropriate to employ the factors identified in other contexts directly when investigating B&Bs. Finally, although majority of the studies have acknowledged the important roles of various factors in influencing consumer satisfaction, most still consider these factors as isolated from each other, without explaining how the processes or paths of various factors affect a customer’s satisfaction. The influencing routes of these factors on a customer’s satisfaction deserve further validation.

Hotel Type Authors Factors Identified

Budget and chain hotels

Mattila and Oneill (2003)

Attentiveness of staff, Guest room cleanliness, Maintenance, Occupancy percentage

Nam et al. (2011) Brand identification, Ideal self-congruence, Lifestyle-congruence, Physical qualities, Staff

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Hotel Type Authors Factors Identified behavior

Deng et al. (2013) Consumption emotions, Customer complaints, Perceived value, Service quality

Ren et al. (2015) Food and beverage service, Guest loyalty programs, Guest rooms, Hotel location, Price and value for money, Security, Staff service

Lee at al. (2016) Customer expectations, Image, Perceived quality, Perceived value

Ren et al. (2016) Aesthetic perception, Location, Staff aspects, Tangible and sensorial experience

Starred hotels

Choi and Chu (2001) Business services, IDD facilities, General amenities, Room qualities, Security, Staff service quality, Value

Kim et al. (2006) Convenience, Information needs, Price benefit, Safety, Service performance and reputation, Technological inclination

Gu et al. (2008) Ancillary services, Bed cleanliness and facilities, Bed comfort, Food and drink, Location and accessibility, Staff performance, Room size and facilities

Lee et al. (2010) Accessibility, Convenience, Safety, Surrounding environment, Tourism attraction, Traffic

Prud’Homme and Raymond (2013)

Charm, Cocooning, Hotel customers’ responsible behaviors, SD orientation

Zhou et al. (2014) Location, Physical setting: food, Physical setting: hotel, Physical setting: room, Staff, Value

Xiang et al. (2014) Core product, Deals, Family friendliness, Hybrid, Staff

Xu and Li (2016) Good location and view, Friendly staff, Nice clean rooms, Good restaurant

Ye and Shu (2016) Access to room, Entrance accessibility, General lodging feature, Information credibility, Luggage and equipment support, Moving convenience, Public area accessibility, Room quality, Room settings, Shower accessibility, Staff attitude and capability

Table 1. Studies exploring factors influencing customer satisfaction with hotels

Methodology

Research Context

The research context of the current study is the ancient town of Pingle, located in Chengdu, in Sichuan Province, one of the largest provinces of China. Located in the southwest of China, Sichuan is less developed compared to the eastern provinces, but is famous for its natural scenery. With rich cultural tourism resources, Sichuan offers fertile ground for B&B tourism development. According to government reports, Sichuan has as many as 31,000 rural homestays (including 3,361 B&Bs) in rural areas, which received over 320 million tourists in 2015 (SichuanOnline 2016). The income from B&Bs in Sichuan was reported as 100.8 billion RMB (about 14.66 billion USD), which makes up 16.23% of the total tourism revenue of this area (SichuanOnline 2016). Sichuan has long been regarded as a popular destination for both domestic and international tourists. Moreover, Chengdu is well connected to all regions of China by direct flights, as well as with most of the capital cities worldwide. Thus, B&Bs located in Pingle ancient town are a suitable choice for our research context.

Data Collection and Analysis

Data was collected from two well-known hotel booking websites: www.qunar.com and www.ctrip.com in China. Qunar.com is a travel search engine and Chinese online travel site that offers domestic and international flights, hotels, tourist resorts, and spot ticket booking services. According to a report

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

from iResearch (2014), the number of monthly active users of Qunar.com exceeded 44 million in September 2014. Ctrip.com is an online ticketing service company, providing more than 90 million members with the full range of travel services, including hotel reservations, ticket booking, travel management etc. In total, 854 reviews for B&Bs located in Pingle posted on these two websites from 13 January 2012 to 28 July 2015 by consumers were collected for analysis.

Data Analysis

The data analysis followed a three-step procedure which is described below.

Content Analysis

An open coding procedure, which involved coding data, splitting or combining categories, then generating new, or dropping existing categories, was adopted to identify the various factors influence customer satisfaction toward B&Bs. A list of factors was obtained based on the phrases and keywords that emerged from the online review data in the content analysis. The relationships between the factors were also identified based on the review data.

ISM Analysis

ISM is an interactive learning process in which a set of varied, but directly related, elements is structured into a comprehensive, systemic model (Sage 1977; Warfield and John 1974). It is a well-established methodology for identifying relationships between specific items that define a problem or an issue (Jharkharia and Shankar 2005). Since its inception, ISM has been applied by a number of researchers in various fields to develop a better understanding of complex systems, such as analyzing vendor selection criteria (Deshmukh and Mandal 1994), exploring the factors affecting flexibility in flexible manufacturing system (Raj 2012), and identifying the barriers to the implementation of total productive maintenance (Attri et al. 2013). In this study, we aim to utilize ISM to explore the relationships between various determinants of customer satisfaction toward B&Bs. Building an ISM involves a number of steps, which are well documented in the literature (Farris and Sage 1975):

Step 1: Defining a set of variables affecting the system (as shown in Table 2);

Step 2: Establishing a contextual relationship between variables (obtained in content analysis);

Step 3: Developing a Reachability Matrix, and checking the matrix for transitivity (as shown in Table 4);

Step 4: Partitioning the Reachability Matrix into different levels (shown in Table 5);

Step 5: Forming a canonical form of matrix (as shown in Table 6);

Step 6: Drawing a directed graph (DIGRAPH) and removing the transitive links;

Step 7: Converting the resultant digraph into an ISM by replacing variable nodes with statements (Figures 2)

Details of the development of the structural models for each of the clusters are set out in the results section.

MICMAC Analysis

Developed by Duperrin and Godet (1973), MICMAC is a systematic analysis tool (Figure 1) for categorizing variables based on hidden and indirect relationships, as well as for assessing the extent to which they influence each other (Hu et al. 2009). Deshmukh and Mandal (1994) claim that the primary goal of MICMAC analysis is to analyze the driver and dependence power of each variable. “Driver power” refers to the degree of influence that one variable has over another, and “dependence power” is defined as the extent to which one variable is influenced by others (Arcade et al. 1999). Based on driver and dependence powers, a 2D driver-dependence diagram can be created, with the horizontal axis representing the extent of dependence and the vertical axis representing the extent of the driver (Arcade et al. 1999). By using MICMAC, the factors can be classified into four clusters, as follows:

Spontaneous factors: factors with weak drive power and weak dependence power. They are relatively disconnected from the system, with which they have only few links.

Dependent factors: factors with weak drive power but strong dependence.

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Linkage factors: factors with strong drive power as well as strong dependence power. These factors are unstable due to the fact that any action on these factors will have an effect on others and also a feedback effect on themselves.

Independent factors: factors with strong drive power but weak dependence power. Factors with relatively strong driving power are “key” variables clustered into the category of independent or linkage factors.

Dependence

I Spontaneous II Dependent

IV Independent III Linkage

Driving

Power

Figure 1. The analysis principle for MICMAC

Data Analysis and Results

Content Analysis Results

The NVivo 10 software was chosen for content analysis. The researchers read each online review and assigned codes to the review data. The relationships between codes were also identified in the reviews. Two coders worked independently on the review data and the independent evaluations were compared. The inter-coder reliability of Krippendorff’s alpha value was calculated, which reached 81.14%, above the 80% acceptance level (Krippendorff 2004). As a result, the 854 online comments produced a total of 104 raw constructs and 43 relationships.

Since there was considerable overlap between the raw constructs, a data reduction process was conducted to consolidate similar constructs (Guo et al. 2010). Constructs that were expressions of the similar underlying idea were combined into one unique construct. For example, “TV” and “water heater” were mapped to the construct “electrical devices”. The consolidation of raw constructs ultimately yielded 46 unique constructs and 29 relationships.

After data reduction, a categorization process was performed on the unique constructs identified to classify them into broader level categories, following the procedure described by Jankowicz and Devi (2004). For instance, “beautiful surrounding environment”, “commercial facilities”, and “noise level” were categorized into one factor dimension, named the “surrounding environment”. In total, 11 factors (including satisfaction) were produced, which are summarized in Table 2 below.

Factors Code Description

B&B Location

S1 This refers to the hotel’s location and nearby facilities (Ren et al. 2015).

Consumption Emotion

S2 This refers to the set of emotional responses elicited during product use or consumption experiences (Havlena and Holbrook 1986).

Expectation Fulfillment

S3 This refers to the extent to which a product’s perceived performance matches the customer’s expectations (Syaqirah and Faizurrahman 2014).

Facility Quality

S4 This refers to the quality of the hotel facilities (Ren et al. 2015).

Perceived Value

S5 This refers to consumer’s overall assessment of the utility of a product or service based on their perceptions of what was received versus what was given (Zeithaml 1988).

Price Level S6 This refers to the total monetary cost to the consumer of purchasing products or service (Jarvenpaa and Todd 1996).

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Room Quality

S7 This refers to the customers’ sensory evaluation of the room quality (Ren et al. 2015).

Satisfaction S8 This refers to the customer’s fulfillment response: a judgment that a product or service provides a pleasurable level of consumption-related fulfillment (Oliver 1997).

Service Quality

S9

This refers to an abstract and elusive construct that has three features unique to services: intangibility, heterogeneity, and inseparability of production and consumption (Parasuraman et al. 1988).

Specialties S10 This refers to the overall architectural style of a B&B hotel that is consistent with the local style of houses, fully embodies the local folk customs, and in harmony with the surrounding environment.

Surrounding Environment

S11 This refers to the supermarkets, restaurants, and other facilities surrounding B&Bs (Kimes and Fitzsimmons 1990).

Table 2. Factors identified in the content analysis

Building the ISM Model

The development of the ISM model begins with identifying a list of variables that are used for modeling. In this study, these variables were obtained from the content analysis of online reviews generated by travelers. In total, 11 variables were obtained in the content analysis and represented by

the symbol 1,2...11iS i , as shown in Table 2. Since the relationships identified in NVivo 10, we

obtained contextual relationships for each factor, as shown in Table 3, in which the cells were populated by 1s and 0s, whereby “1” indicated the relationship and “0” indicated otherwise. Based on the Adjacency Matrix, Reachability Matrix (Tables 4) was calculated using Matlab. The driving power of a particular factor is the total number of factors (including itself) that it influences. The dependence is the total number of factors (including itself) that may help in influencing its development (Qureshi 2013). These driving power and dependency values will be used in the MICMAC analysis, where the factors will be classified into four groups of autonomous, dependent, linkage, and independent factors (Faisal et al. 2006).

S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S1 0 1 1 0 1 1 0 0 0 0 1 S2 0 0 1 0 1 0 0 1 0 0 0 S3 0 0 0 0 0 0 0 1 0 0 0 S4 0 1 1 0 1 0 1 1 0 0 0 S5 0 0 0 0 0 0 0 1 0 0 0 S6 0 1 0 0 1 0 0 0 0 0 0 S7 0 1 1 0 1 0 0 1 0 0 0 S8 0 0 0 0 0 0 0 0 0 0 0 S9 0 1 1 0 1 0 0 0 0 0 0 S10 0 1 1 0 0 0 0 0 0 0 0 S11 0 1 0 0 0 0 0 1 0 0 0

Table 3. Adjacency matrix

S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 Driving Power

S1 1 1 1 0 1 1 0 1 0 0 1 7

S2 0 1 1 0 1 0 0 1 0 0 0 4

S3 0 0 1 0 0 0 0 1 0 0 0 2

S4 0 1 1 1 1 0 1 1 0 0 0 6

S5 0 0 0 0 1 0 0 1 0 0 0 2

S6 0 1 1 0 1 1 0 1 0 0 0 5

S7 0 1 1 0 1 0 1 1 0 0 0 5

S8 0 0 0 0 0 0 0 1 0 0 0 1

S9 0 1 1 0 1 0 0 1 1 0 0 5

S10 0 1 1 0 1 0 0 1 0 1 0 5

S11 0 1 1 0 1 0 0 1 0 0 1 5

Dependence 1 8 9 1 9 2 2 11 1 1 2 47

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Table 4. Reachability matrix

In the reachability matrix, every variable has a reachability set and an antecedent set (Warfield and John 1974). We can derive the intersection of the reachability set and antecedent set for every variable. The factors for which the reachability and the intersection sets are the same that occupy the top level in the ISM hierarchy, which will not help achieve any other variable above their own level (Bhattacharya and Momaya 2009). Once the top-level factor is identified, it is removed from the subsequent iteration (Bhattacharya and Momaya 2009). Table 5 below shows the result of level partitions. The identified levels aid in building the digraph and the ISM model. And a canonical matrix is developed by clustering factors at the same level across the rows and columns of the final reachability matrix (Agarwal et al. 2007). The canonical matrix helps in the generation of the structural model. The canonical matrix is shown in Table 6 below.

Factor Reachability set Antecedent set Intersection set Level S1 1,2,3,5,6,11 1 1 L5

S2 2,3,5 1,2,4,6,7,9,10,11 2 L3

S3 3 1,2,3,4,6,7,9,10,11 3 L2 S4 2,3,4,5,7 4 4 L5

S5 5 1,2,4,5,6,7,9,10,11 5 L2 S6 2,3,5,6 1,6 6 L4 S7 2,3,5,7 4,7 7 L4 S8 8 1,2,3,4,5,6,7,8,9,10,11 8 L1 S9 2,3,5,9 9 9 L4 S10 2,3,5,10 10 10 L4 S11 2,3,5,11 1,11 11 L4

Table 5 Iterations

S8 S3 S5 S2 S6 S7 S9 S10 S11 S1 S4 Driving Power S8 1 0 0 0 0 0 0 0 0 0 0 1 S3 1 1 0 0 0 0 0 0 0 0 0 2 S5 1 0 1 0 0 0 0 0 0 0 0 2 S2 1 1 1 1 0 0 0 0 0 0 0 4 S6 1 1 1 1 1 0 0 0 0 0 0 5 S7 1 1 1 1 0 1 0 0 0 0 0 5 S9 1 1 1 1 0 0 1 0 0 0 0 5 S10 1 1 1 1 0 0 0 1 0 0 0 5 S11 1 1 1 1 0 0 0 0 1 0 0 5 S1 1 1 1 1 1 0 0 0 1 1 0 7 S4 1 1 1 1 0 1 0 0 0 0 1 6

Dependence 11 9 9 8 2 2 1 1 2 1 1 47 Table 6 Canonical matrix

Figure 2 shows the structural model for this study. The digraph illustrates the direct relations among all factors, with arrows indicating the direction of each impact. The results show that these factors are organized into five levels in the hierarchy model. “Satisfaction” is the target of the system, and is located in the top level, which directly depends on “expectation fulfillment” and “perceived value”. “Consumption emotion” is the only factor in the third level. It directly depends on the lower-level variables of “price level”, “room quality”, “service quality”, “surrounding environment”, and “specialties”. “Consumption emotion” transforms influences from lower levels to higher levels. The bottom-level variables are “facility quality” and “B&B’s location”, which are considered to be strong enablers of “price level”, “room quality”, and “surrounding environment”.

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

L5

L4

L3

L2

S8 Satisfaction

S3

Expectation Fulfillment

S5

Perceived Value

S2

Consumption Emotion

S6

Price LevelS7

Room Quality

S11

Surrounding Environment

S4

Facility Quality

S10

Specialties

L1

S1

B&Bs Location

S9

Service Quality

Figure 2. ISM-based model for determinants of customer satisfaction toward B&Bs

MICMAC Analysis

The purpose of the MICMAC method in this study is to help classify these determinants of customer satisfaction. By summing up the diver power and dependence and then dividing them by 11 (the total number of variables considered), we obtained an average of 4.27 for both parameters. This value was used to separate the driver and dependence diagram to the four quadrants shown in Figure 3. Consequently, these factors were classified to four clusters. The driver-dependence diagram we obtained shows that there are no variables in spontaneous clusters, meaning that no variable is totally disconnected within our system. Thus, we should consider all factors. The second cluster of dependent variables includes “expectation fulfillment”, “perceived value”, “consumption emotion”, and “customer satisfaction”. They have the greatest dependence and the least driving power, lying at the top level of ISM hierarchy. Because of their high dependence, they require all of the other variables to advance them in achieving the goal of the system. Although there is no variable present in the linkage cluster, it is worthwhile mentioning that the factor “consumption emotion” falls on the border of the linkage cluster. The independent cluster includes “price level”, “room quality”, “service quality”, “specialties”, “surrounding environment”, “location”, and “facility quality”, which are located in the bottom levels of the ISM model. These variables have high driving power and low dependence, indicating that they have a strong influence on others in the dependent cluster. Generally, they can be regarded as the base of the management strategy by decision-makers.

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

2

0

0 4 6 8 10 12

2

4

6

8

10

12

4.27

4.27

S1

Dependence

Driving

Power S2

S3

S5

S8

S7

S6

S11S10

S9

S4

Figure 3. Driver power and dependence diagram

Discussion

Using content analysis based on user-generated reviews, our study identified ten factors important in driving consumer satisfaction with B&Bs. These factors were organized into a five-level hierarchical model using ISM. Comparing our results with those in the literature, we have a few interesting findings. Firstly, the bottom-level factors: “location” and “facility quality” serve as two strong drivers of customer satisfaction. “Location” directly leads to “room quality”, while “facility quality” directly leads to “price level” and “surrounding environment”. These results are in line with those from previous studies (e.g., Lee et al. 2010; Ren et al. 2016; Ren et al. 2015; Zhou et al. 2014) demonstrating that location plays an important role in driving consumer satisfaction in both the budget and starred hotel contexts. Moreover, the “facility quality” of B&Bs is also far worse than that of the budget and chain hotels. However, we found that the “facility quality” was mentioned in 45.67% of the reviews, indicating that consumers still have high expectations of the facility quality of B&Bs. Thus, if B&B operators slightly improve the quality of their facilities, consumer perceptions and satisfaction will be greatly improved.

Secondly, “price level”, “room quality”, “service quality”, “surrounding environment”, and “specialties”, all in the lower level of the model, are the basic factors for customer satisfaction with B&Bs. Improved customer satisfaction can be achieved indirectly by improving these factors, since they had high driver power. Among these factors, “room quality” and “service quality” are more important as they are mentioned in 41.1% and 41.15% of the reviews respectively. Both of these factors have been frequently examined as determinants of satisfaction in the budget hotel or star hotel context (e.g., Lee et al. 2010; Mattila and Oneill 2003; Nam et al. 2011; Ren et al. 2016; Syaqirah and Faizurrahman 2014). Our results indicate that they can be applied to the B&B context as well. Moreover, we found that “specialties” is unique to the B&B context. “Specialties” means that the overall architectural style of the B&B is consistent with the local house style, fully embodies the local folk customs, and is in harmony with the surrounding environment (Lin et al. 2012). This includes both the interior decoration and architectural design of the B&B, which will satisfy customers’ demand for local cultural characteristics. Most B&Bs are located in cultural town or villages. Thus, compared to general hotels, B&Bs not only have dominant locations close to scenic attractions, but also provide a special accommodation experience with local characteristics.

Thirdly, “consumption emotion”, which refers to the set of emotional responses elicited during product usage or consumption experiences (Havlena and Holbrook 1986), comes third in our hierarchy. It falls in the dependent cluster of the MICMAC analysis. However, it is close to the linkage cluster and is indeed imperative in translating independent factors to effective influence of the dependent factors. From the hierarchical model it is evident that “consumption emotion” plays the most important role in influencing customer satisfaction, since all independent variables go through it to influence consumer satisfaction. Only when consumers experience favorable emotions throughout their experience, will they feel satisfied with that experience (Deng et al. 2013).

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Fourthly, “expectation fulfillment” and “perceived value” are two psychological factors that can directly lead to satisfaction. These two factors can only be achieved directly through “consumption emotion”. While previous studies mainly focused on attribute-level factors such as service quality, location, and room quality (e.g., Gu et al. 2008; Xiang et al. 2014; Xu and Li 2016), we extend their findings to uncover the higher level psychological factors that directly correspond with satisfaction. Thus, B&B operators should consider consumers’ psychological responses as well.

Finally, we identified the interrelationships and influencing path between these factors. As shown in Figure 2, B&B’s location and facility quality, which were located at the bottom of the model, have direct impact on price level, surrounding environment and room quality in the next hierarchy. The factor of consumption emotion is imperative for translating lower levels’ factors into effective influence towards customer satisfaction. Any shortcomings caused by basic factors could have a negative impact on fulfilling customers’ emotion. There were two factors at the top of the ISM model with highest dependence, which means that they were influenced by lower level factors. Any action on any other factors has an impact on them.

Implications

From the theoretical perspective, this study demonstrated the usefulness of big data analytics in identifying determinants of hotel guest satisfaction using consumer generated content readily available on the Internet. While our findings were based upon data generated on a specific website and during a specific time period, they reflected the way consumers “talk about” their experiences in online reviews. Our analysis revealed the opinion differentiations in relation to influencing factors expressed in the online reviews. Moreover, this study is among the first to systematically explore the determinants of customer satisfaction in the B&B context and uncover the influencing process or the paths of various factors influencing customers’ satisfaction. It offers a list of determinants of satisfaction that has little researcher bias, and thus can be used for the development of questionnaires for customer satisfaction studies.

This study also has practical implications for B&B operators making marketing strategies. Compared to the traditional budget hotels and starred hotels, B&B customers are more concerned with the emotional experience associated with B&B accommodation. First, hotel managers should fully utilize eWOM by applying opinion mining towards customer online reviews in the hospitality industry of today. Second, B&B operators need to choose a convenient location where customers not only have easy access to affordable dining facilities, supermarkets, and public transportation, but also can enjoy the local scenic attractions easily. Moreover, operators could provide a warm human experience that is different to that of traditional hotel accommodation.

Limitations and Future Direction

The limitation of this study is in its research scope and generalizability. The current study only examined the B&Bs in Pingle ancient town, located in Chengdu, China. It remains to be verified whether the results apply to B&Bs in other countries. Future research may expand this research scope and verify the findings of this study by using review data for B&Bs in other cultural contexts. Online comments from other websites, such as www.booking.com and www.agoda.com, can be considered as data sources and used to triangulate the findings of this study.

References

Agarwal, A., Shankar, R., and Tiwari, M. K. 2007. "Modeling Agility of Supply Chain," Industrial Marketing Management (36:4), pp. 443-457.

Anderson, E. W., and Sullivan, M. W. 1993. "The Antecedents and Consequences of Customer Satisfaction for Firms," Marketing science (12:2), pp. 125-143.

Arcade, J., Godet, M., Meunier, F., and Roubelat, F. 1999. "Structural Analysis with the Micmac Method & Actor's Strategy with Mactor Method," Futures Research Methodology, American Council for the United Nations University: The Millennium Project), pp. 1-69.

Attri, R., Grover, S., Dev, N., and Kumar, D. 2013. "Analysis of Barriers of Total Productive Maintenance (Tpm)," International Journal of System Assurance Engineering & Management (4:4), pp. 365-377.

Barsky, J. D. 1992. "Customer Satisfaction in the Hotel Industry: Meaning and Measurement," Journal of Hospitality & Tourism Research (16:1), pp. 51-73.

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Bhattacharya, S., and Momaya, K. 2009. "Interpretive Structural Modeling of Growth Enablers in Construction Companies," Singapore Management Review (31:1).

Chen, J. L. 2015. "The Impact of Bed and Breakfast Atmosphere, Customer Experience, and Customer Value on Customer Voluntary Performance: A Survey in Taiwan," Asia Pacific Journal of Tourism Research (20:5), pp. 541-562.

Chen, L. C., Lin, S. P., and Kuo, C. M. 2013. "Rural Tourism: Marketing Strategies for the Bed and Breakfast Industry in Taiwan," International Journal of Hospitality Management (32:1), pp. 278-286.

Choi, T. Y., and Chu, R. 2001. "Determinants of Hotel Guests’ Satisfaction and Repeat Patronage in the Hong Kong Hotel Industry," International Journal of Hospitality Management (20:3), pp. 277-297.

Deng, W. J., Yeh, M. L., and Sung, M. L. 2013. "A Customer Satisfaction Index Model for International Tourist Hotels: Integrating Consumption Emotions into the American Customer Satisfaction Index," International Journal of Hospitality Management (35), pp. 133-140.

Deshmukh, S. G., and Mandal, A. 1994. "Vendor Selection Using Interpretive Structural Modelling (Ism)," International Journal of Operations & Production Management (14:6), pp. 52-59.

Duperrin, J. C., and Godet, M. 1973. "Méthode De Hiérarchisation Des léments D'un Système : Essai De Prospective Du Système De L'énergie Nucléaire Dans Son Contexte Sociétal,").

Faisal, M. N., Banwet, D. K., and Shankar, R. 2006. "Supply Chain Risk Mitigation: Modeling the Enablers," Business Process Management Journal (12:4), pp. 535-552.

Farris, D. R., and Sage, A. P. 1975. "On the Use of Interpretive Structural Modeling for Worth

Assessment ☆," Computers & Electrical Engineering (2:2-3), pp. 149-174. Grönroos, C. 1983. "Strategic Management and Marketing in Service Sector," Journal of Business

Research). Gu, H. M., Ryan, C., and Wang, Y. C. 2008. "Chinese Clientele at Chinese Hotels - Preferences and

Satisfaction," International Journal of Hospitality Management (27:3), pp. 337-345. Guo, Z., Tan, F. B., and Cheung, K. 2010. "Students' Uses and Gratifications for Using Computer-

Mediated Communication Media in Learning Contexts," Communications of the Association for Information Systems (27:1), pp. 339-378.

Havlena, W. J., and Holbrook, M. B. 1986. "The Varieties of Consumption Experience: Comparing Two Typologies of Emotion in Consumer Behavior," Journal of Consumer Research (13:3), pp. 394-404.

Hu, H. Y., ShaoI, Chiu, and Yen, T. M. 2009. "Modified Ipa for Order-Winner Criteria Improvement: A Micmac Approach," Journal of Applied Sciences (9:21).

Iresearch. 2014. "The Introduction of Qunar.Com." from http://www.qunar.com/site/zh/Qunar.in.China_1.2.shtml

Jankowicz, and Devi. 2004. "The Easy Guide to Repertory Grids,"). Jarvenpaa, S. L., and Todd, P. A. 1996. "Consumer Reactions to Electronic Shopping on the World

Wide Web," International Journal of Electronic Commerce (1:2), pp. 59-88. Jharkharia, S., and Shankar, R. 2005. "It-Enablement of Supply Chains: Understanding the Barriers,"

Journal of Enterprise Information Management (18:1), pp. 11-27. Kim, J., and Hardin, A. 2010. "The Impact of Virtual Worlds on Word-of-Mouth: Improving Social

Networking and Servicescape in the Hospitality Industry," Journal of Hospitality Marketing & Management (19:7), pp. 735-753.

Kim, W. G., Ma, X. J., and Kim, D. J. 2006. "Determinants of Chinese Hotel Customers' E-Satisfaction and Purchase Intentions," Tourism Management (27:5), pp. 890-900.

Kimes, S. E., and Fitzsimmons, J. A. 1990. "Selecting Profitable Hotel Sites at La Quinta Motor Inns," Interfaces (20:2), pp. 12-20.

Krippendorff, K. 2004. "Reliability in Content Analysis," Human Communication Research (30:3), pp. 411-433.

Kuo, Y. H., Chao, H. C., and Weng, W. T. 2014. "A Comparison Study of Housing-Performance between B&B and Hotel Industry in Taiwan," Procedia - Social and Behavioral Sciences (144), pp. 79-90.

Lee, K.-W., Kim, H.-b., Kim, H.-S., and Lee, D.-S. 2010. "The Determinants of Factors in Fit Guests' Perception of Hotel Location," Journal of Hospitality and Tourism Management (17:1), pp. 167-174.

Lin, C. F., Hueijen, S., and Wu, M. Y. 2012. "Exploring the Relationships between Tourist Perceptions and the Service Quality of B&Bs," Journal of Travel & Tourism Marketing (29:6), pp. 585-598.

Lynch, P. A. 2005. "The Commercial Home Enterprise and Host: A United Kingdom Perspective," International Journal of Hospitality Management (24:4), pp. 533-553.

Web Opinion Mining For B&B Customer Satisfaction

Twenty First Pacific Asia Conference on Information Systems, Langkawi 2017

Mattila, A. S., and Oneill, J. W. 2003. "Relationships between Hotel Room Pricing, Occupancy, and Guest Satisfaction: A Longitudinal Vase of a Midscale Hotel in the United States," Journal of Hospitality & Tourism Research (27:3), pp. 328-341.

Nam, J., Ekinci, Y., and Whyatt, G. 2011. "Brand Equity, Brand Loyalty and Consumer Satisfaction," Annals of Tourism Research (38:3), pp. 1009-1030.

Nuntsu, N., Tassiopoulos, D., and Haydam, N. 2004. "The Bed and Breakfast Market of Buffalo City (Bc), South Africa: Present Status, Constraints and Success Factors," Tourism Management (25:4), pp. 515-522.

Oliver, R. L. 1997. "Satisfaction: A Behavioral Perspective on the Consumer," Asia Pacific Journal of Management:2), pp. 285-286.

Parasuraman, A., Zeithaml, V. A., and Berry, L. L. 1988. "‘Servqual: A Multiple-Item Scale for Measuring Consumer Perceptions of Service Quality’," Journal of Retailing (64(1):1), pp. 12-40.

Prud’Homme, B., and Raymond, L. 2013. "Sustainable Development Practices in the Hospitality Industry: An Empirical Study of Their Impact on Customer Satisfaction and Intentions," International Journal of Hospitality Management (34:6), pp. 116-126.

Qureshi, M. N. 2013. "An Integrated Model to Identify and Classify the Key Criteria and Their Role in the Assessment of 3pl Services Providers," Asia Pacific Journal of Marketing & Logistics (20:20), pp. 227-249.

Raj, T. 2012. "Modelling the Factors Affecting Flexibility in Fms," International Journal of Industrial & Systems Engineering (11:4), pp. 350-374.

Ren, L., Qiu, H., Wang, P., and Lin, P. M. C. 2016. "Exploring Customer Experience with Budget Hotels: Dimensionality and Satisfaction," International Journal of Hospitality Management (52), pp. 13-23.

Ren, L. P., Zhang, H. Q., and HaoBin, Y. 2015. "Understanding Customer Satisfaction with Budget Hotels through Online Comments: Evidence from Home Inns in China," Journal of Quality Assurance in Hospitality & Tourism (16:1), pp. 45-62.

Sage, A. P. 1977. Methodology for Large Scale Systems. McGraw-Hill. SichuanOnline. 2016. "The Birthplace of Chinese Agritainment." from

http://sichuan.scol.com.cn/ggxw/201601/54263201.html Suzhouminsu. 2016. "Research Report About the Development of Chinese B&B in 2016." from

http://www.suzhouminsu.com/word/wenzhang_081.html Syaqirah, Z. N., and Faizurrahman, Z. P. 2014. "Managing Customer Retention of Hotel Industry in

Malaysia ☆," Procedia - Social and Behavioral Sciences (130:1-3), pp. 379-389. Turner, S. A. 2011. "Study: B&Bs Could See Business Boom." from http://www.HotelNewsNow.com Warfield, and John, N. 1974. "Developing Interconnection Matrices in Structural Modeling," IEEE

Transactions on Systems Man & Cybernetics (SMC-4:1), pp. 81-87. Xiang, Z., Schwartz, Z., Gerdes, J. H., and Uysal, M. 2014. "What Can Big Data and Text Analytics Tell

Us About Hotel Guest Experience and Satisfaction?," International Journal of Hospitality Management (44:44), pp. 120-130.

Xu, X., and Li, Y. 2016. "The Antecedents of Customer Satisfaction and Dissatisfaction toward Various Types of Hotels: A Text Mining Approach," International Journal of Hospitality Management (55), pp. 57-69.

Ye, Z., and Shu, T. C. 2016. "Dimensions of Lodging Guest Satisfaction among Guests with Mobility Challenges: A Mixed-Method Analysis of Web-Based Texts," Tourism Management (53), pp. 13–27.

Zane, B. 1997. "The B&B Guest: A Comprehensive View," Cornell Hotel and Restaurant Administration Quarterly (38:4), pp. 69-75.

Zeithaml, V. A. 1988. "Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence," Journal of Marketing (52:3), pp. 2-22.

Zhou, L., Ye, S., Pearce, P. L., and Wu, M. Y. 2014. "Refreshing Hotel Satisfaction Studies by Reconfiguring Customer Review Data," International Journal of Hospitality Management (38:38), pp. 1-10.