Accepted Manuscript

Determinants of behavioral intention to use the personalized

location-based mobile tourism application: An empirical study by
integrating TAM with ISSM

Chia-Chen Chen, Jia-Lun Tsai

PII: S0167-739X(17)30253-4
DOI: [Link]
Reference: FUTURE 3346

To appear in: Future Generation Computer Systems

Received date: 31 August 2016

Revised date: 4 February 2017
Accepted date: 15 February 2017

Please cite this article as: C.-C. Chen, J.-L. Tsai, Determinants of behavioral intention to use
the personalized location-based mobile tourism application: An empirical study by integrating
TAM with ISSM, Future Generation Computer Systems (2017),
[Link]

This is a PDF file of an unedited manuscript that has been accepted for publication. As a
service to our customers we are providing this early version of the manuscript. The manuscript
will undergo copyediting, typesetting, and review of the resulting proof before it is published in
its final form. Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
 Determinants of behavioral intention to use the Personalized
Location-based Mobile Tourism Application: An empirical study by
integrating TAM with ISSM

Chia-Chen Chen*, Jia-Lun Tsai

emily@[Link], z85062@[Link]

Department of Management Information Systems, National Chung

Hsing University, Taichung 402, Taiwan, R.O.C

Correspondence: Chia-Chen Chen, Associate Professor

Address: No.250, Kuo Kuang Rd., Taichung City 402, Taiwan, R.O.C.

Tel: +886-4-22840864 ext 667

Fax: +886-4-22857173

E-mail: emily@[Link]

1
 Determinants of behavioral intention to use the Personalized
Location-based Mobile Tourism Application: An empirical study by
integrating TAM with ISSM

Abstract

With the advent of the two-day weekend and improvements in the public
transit system, people have begun to focus on leisure activities. When the YouBike
public bicycle system was installed in the city of Taichung, Taiwan, it created a
convenient transportation system network that was set up perfectly for a tremendous
impact on the local tourism industry. This has happened in parallel with the
development and proliferation of smartphones and wireless networks. The functions
of mobile applications (“apps”) have become more powerful over time, allowing
people to access travel information and share their experiences almost
instantaneously. Since a smartphone’s positioning system can be used to provide
more personalized information and services, the development trend is heading
toward location-based services (LBSs) that can bring the app’s functionality closer
to the needs of the user.
This study develops a personalized location-based mobile tourism application
(PLMTA) for travel planning. The PLMTA combines hybrid filtering technology
with the ant colony optimization (ACO) algorithm to make more efficient
customized tourism recommendations. It allows users to more effectively search
through travel information and arrange their trip. This study also integrates the
technology acceptance model (TAM) and the information system success model
(ISSM) to present a research model that explores users’ intention to use the PLMTA.
The questionnaire survey method is used to collect our data, and the hypotheses are
tested via structural equation modeling (SEM). The results show that information
quality, perceived ease of use, and perceived usefulness significantly affect the
intention to use PLMTA, while information quality and perceived convenience are
found to have an influence on perceived usefulness. Information quality, system
quality, and perceived convenience are found to significantly affect perceived ease
of use, which consequently affects the intention to use the system.

Keywords: Location-based Services, Ant Colony Optimization, Technology

Acceptance Model, Information System Success Model

2
1. Introduction

According to a survey by eMarketer (2014), the number of smartphone users

reached 17.5 billion, globally, in 2014, having just passed 10 billion mark in 2012

[1]. In addition, the smartphone penetration rate is expected to be close to 50% in

2017 [1]. The booming development of smartphones and the increasing

sophistication of positioning systems and navigation functions have spurred the

constant development of new mobile apps. The smartphone’s powerful

functionalities can be combined with location-based services (LBSs) which use the

mobile phone’s positioning function to capture the user’s location in order to

provide personalized services suitable to that location for greater convenience. The

combinations of these smart technology applications and location-based services

can be quite diverse. For example, retailers use indoor navigation technology to

provide location-based services, using mobile “push” notifications to deliver ads,

thus providing appropriate, personalized marketing based on the consumer’s

location to enhance the quality of the customer experience. This kind of technology

is set to become an important development trend in the future (Market Intelligence

& Consulting Institute, 2015). Combining LBS and mobile phone apps has already

become a global trend, providing convenient geographic information and

information that relates directly to the user’s specific location. Examples include

store promotions, exhibition activity information, coupons and mobile education. As

location-based services have matured, Click-and-Mortar has become both a

possibility and a reality (TELDAP e-Newsletter, 2010).

In response to advances in information technology, new information systems

are constantly being developed, and an increasing number of apps has accompanied

the flourishing of smartphones and network services. The tourism recommender

3
system is one such app. From the original, web-based-only system interface to

mobile apps, users now have a variety of flexible options. The rapid development of

science and technology has allowed tourism recommender systems to increase their

functionality with new artificial intelligence technology such as multi-agent

technology, optimization algorithms, cluster analysis, and so on. Recommender

systems have become more complete and accurate, with interface designs that are

more user-friendly [2].

While many applications combine LBS with mobile apps, the most common

type combines the phone’s positioning system with a web-based electronic map

(e.g., Google Map, Yahoo Local, UrMap). Such applications provide an interface

that uses e-maps and geological information system functions to offer users

personalized information and services. Combining web-based electronic maps with

location-based services allows for recommendations that are more accurate and

better aligned with users’ current needs. More relevant functions such as route

planning, time estimation, and online coupons also become available. The map

interface also allows users to be apprised of the distance between points at a glance,

giving them more relevant geographical information.

As environmental awareness has increased, global warming and energy

shortages have become increasingly apparent. To improve air quality and noise

issues, many countries have begun to actively promote green vehicles. The bicycle

sharing system is one such method [3]. The Taipei City Government worked in

cooperation with Taiwanese bicycle company Giant to promote the YouBike public

bicycle sharing system in Taipei, Taiwan. YouBike uses advanced radio frequency

identification (RFID) technology, for increased user convenience. Chung (2010)

found that the level of satisfaction users feel regarding YouBike’s overall service is
4
good, but attention must be paid to factors such as user perceived quality (both

system and service quality) and corporate image (brand awareness) in order to build

customer loyalty, which enhances users’ willingness to continue to use the YouBike

system [4]. In view of YouBike’s success in Taipei, the city government of Taichung,

in central Taiwan, began building a public bicycle sharing system called “iBike” in

2014. The government hopes to construct a comprehensive and convenient public

transportation system network for short distance transport, improving its citizens’

quality of life and promoting tourism.

Tourism recommender systems rarely use the mode of transportation as the

main consideration for system design. However, traffic problems are a major

problem for foreign tourists. Personalized recommendations are for individual users,

providing one-on-one recommendation results [5, 6]. Many studies indicate that

tourism recommender systems that offer personalized services are more likely to

attract user interest and can more effectively assist users to plan their trips, thereby

enhancing user satisfaction [7, 8, 9, 10]. Therefore, this study designs a personalized

location-based mobile tourism application (PLMTA). This PLMTA uses hybrid

filtering technology to gather tourism information more efficiently. The system also

adopts an ant colony optimization (ACO) algorithm to customize tourism

recommendations with location-relevant featured transportation to make trip

planning easier. Thus, this study designs a set of tour planning functions based on

the iBike system in metropolitan Taichung. In addition to tourism recommendation

functions, the app participates in the promotion and development of Taichung’s

iBike system. It includes iBike stations and neighboring attractions as tour planning

options, helping tourists to plan their Taichung trip more flexibly and conveniently.

Lastly, this study integrates the technology acceptance model (TAM) and the

5
information system success model (ISSM) to construct a research model to

investigate users’ intention to use the PLMTA. The study aims to understand how to

design and develop a mobile tourism application as a personalized recommender

system that is attractive to users and can improve daily life through “smart living.”

2. Literature Review

2.1. Recommender Systems and Related Technologies

A recommender system is defined as one that provides “personal data or

preferences as input data, after system processing and cross-check, to provide

recommendation results to the appropriate recipients” [6]. Algorithm improvements

and breakthroughs allow recommender systems to target specific groups. There are

two main technical applications in a recommender system: content filtering (CB)

and collaborative filtering (CF). Since these two technologies have different issues,

their applications are limited. Over the years, numerous attempts have been made to

overcome these limitations, and many different recommender system technologies

have been developed. The four introduced below are the most common [11, 12].

(1) Content Filtering (CB)

Content filtering, analyzes user preferences, past selection history records, and

information regarding chosen products in order to recommend products that are

similar to customers’ preferences. In the initial stage, a questionnaire is used to

gather the preferences of new users. Cross-referencing and filtering produce a list of

similar products to be recommended. Recommendation results based on content

filtering are generally better than those based on collaborative filtering. However,

the recommendation results tend to be too similar, offering users only a limited

selection. Moreover, it is impossible to recommend products to brand new users or

6
to those who rarely fill out questionnaires.

(2) Collaborative Filtering (CF)

Collaborative filtering involves collecting the preferences of groups of users

who have similar interests or the same experience. This kind of filtering can be

further divided into two types: user-based and item-based. User-based collaborative

filtering analyzes all users’ demand and preference for commodities, and conducts a

similarity cross-reference to find other users whose preferences are similar to those

of the current user as the basis for a recommendation. On the other hand, Sarwar et

al. (1999) concluded that, in order to attract user attention, an item must be similar

to those that have high ratings [13]. Therefore, item-based collaborative filtering is

based on the calculated similarities between items instead of the calculated

similarities between users. The problem with collaborative filtering, however, is that

new products are ignored by the filter because they have no usage records.

Therefore, it is not appropriate to use the collaborative filtering recommendation

method for new products.

(3) Hybrid Filtering

Since content filtering and collaborative filtering recommendations have

distinctly different limitations, many recommender systems overcome these by

using a hybrid method which combines the two. There are three common hybrid

approaches: one is to execute content filtering and collaborative filtering separately,

and then combine the prediction results; another is to integrate some features of

content filtering into the collaborative filtering; and the last is to integrate some

features of collaborative filtering into the content filtering.

(4) Demographic Filtering (DF)

Demographic filtering employs the user’s personal attributes (e.g., gender,

7
occupation, age) to find other users with similar attributes. This kind of filtering

classifies users with similar features into groups and tracks the preferences and

behavior of other users in the same category.

The range of recommender system applications has become quite extensive.

As the popularity of smartphones has risen, mobile app recommender systems that

use positioning services have become a major trend. Based on the geographical

location acquired from the smartphone’s positioning system, the filters find more

appropriate results which are then recommended to users. When the recommended

results and services align with users’ expectations, users are more satisfied with the

app. Recommender apps commonly feature neighboring restaurants, gas stations,

parking lots, and public transportation, all of which are highly relevant to people’s

daily lives. As information technology continues to evolve, recommender systems

can be applied to more fields and other aspects of daily living. When combined with

the real-time and ubiquity of mobile apps, recommender systems can increase

convenience and improve people’s quality of life.

Since each filtering approach produces different results, recommender system

that adopt only one filtering technology are more limited in their ability to match the

usage context and fulfill users’ requirements. Therefore, this study adopts hybrid

filtering as the main filtering approach in the PLMTA.

2.2. Tourism Recommender Systems

In response to the rapid evolution of smartphones and mobile networks, users

are no longer constrained by time and space. A tour recommender system works in

combination with the smartphone’s positioning system to collect tour information

that more closely meets the user’s needs, providing a higher quality LBS. It is more

flexible to use and the operation is more immediate, so overall efficiency is

8
enhanced. Borras et al. (2014) organized and analyzed tourism recommender

systems from 2008 to 2013 and found that, in comparison to collaborative and

demographic filtering, content filtering technology was adopted by more tourism

recommender systems [2]. Tourism recommender systems that use only a single

filtering technology are now in the minority. Hybrid filtering that combines two or

more technologies is now used by 53% of such systems, resolving a diverse set of

problems.

Borras et al. (2014) also pointed out that tourism recommender system

functions can be generally divided into four categories: recommended tour

destinations and packages, recommended attractions and ratings, trip planning, and

social networking [2]. The main function of most tourism recommender systems is

to recommend and rate attractions. Fewer systems have trip planning as their main

function. Some recommender systems work in combination with the information

sharing functions of social networks. Few tourism recommender systems provide

complete tour destination information and tour packages as their main function.

Batet et al. (2012) proposed the Turist@ system based on multi-agent

technology to give personalized tour attraction recommendations more effectively

[14]. Yang and Hwang (2013) proposed the ITravel recommender system in a

mobile peer-to-peer environment [15], allowing users to discuss the reviews of

attractions via peer-to-peer communication. The map interface lists the user’s

nearby friends so that the user can ask their opinions regarding trip plans and

adjustments. Zhou et al. (2014) developed the SoLoMo system using the k-nearest

neighbors algorithm to consider both geographical distance and social distance as

the basis for its recommendations [16]. Chiang and Huang (2015) designed an

algorithm that takes time, user preferences and other factors into consideration to

recommend the most appropriate itinerary [17]. He et al. (2016) developed the
9
SocoTraveler model to leverage the individual travel history and social influence of

co-travelers to analyze personal interests and then provide potential

recommendations for users [18]. Cenamor et al. (2017) proposed the PlanTour

system to provide personalized tourist plans by collecting human-generated

information from travel social network sites in the “minube” travel app [19].

2.3. Location-Based Services

Location-based services provide the most appropriate services and information

that can be offered at the users’ current location after confirming the user’s

geographical position via the combined functionality of a mobile device, mobile

network, and the global positioning system [20]. The Open Geospatial Consortium

(2005) decreed that servicing mobile network users by combining wireless network

service with geographical information shall be referred to as “location-based service”

[21]. As the penetration rate of smartphones is high and with the rapid changes in

mobile networks, new mobile apps are being developed constantly and continuously.

Using geographic location information, LBS mobile apps provide more satisfying

personalized services.

Statistics gathered by market research website eMarketer (2014) indicate that

approximately 60% of mobile phone users use an LBS mobile app at least once a

month, and this usage trend has increased year after year [1]. LBS applications are

becoming increasingly diverse, often used in searches for nearby shops and

restaurants. Samsioe & Samsioe (2002) pointed out that applying LBS to

e-commerce provides users with the best service quality by offering specific

services based on the geographical location obtained by the positioning system of

the mobile device [22]. Location-based services are also widely used in mobile

10
commerce, in which the most appropriate services and advertisements are pushed to

potential consumers according to their needs and current location [23]. Thus, the

major trend is to combine LBS and electronic maps to allow the user to have a clear

concept of travel time and distance. Applying this concept to a tourism

recommender system can facilitate the collection of tour information and tour

planning.

Yang and Wang (2009) integrated GPS and web2.0 to create a set of mobile

location-based information recommender systems to provide personalized

recommendations [24]. Their system used the dynamic 3D text cloud to make data

collection more efficient and provide users with dynamic recommendations based

on synergy-added information. Yu et al. (2009) built an intelligent mashup

recommender system based on the concept of ontology [25]. Once users revealed

their tourism needs and preferences, the system took demand conditions and past

experience into consideration to recommend appropriate tourism information.

Husain and Dih (2012) adopted the TF-IDF (Term Frequency-Inverse Document

Frequency) algorithm to filter the content and build a personalized location-based

tourism recommender system [26]. Gavalas et al. (2014) adopted the systematic

approach to review the state-of-the-art in recommender systems, especially for the

classification of mobile tourism applications. They determined that a massive

amount of information can be gathered from LBSs, web technologies and social

networking to provide highly accurate and effective tourism recommendations [27].

2.4. Ant Colony Optimization (ACO)

The ant algorithm was first proposed by Dorigo in 1996, originally intended to

solve the problems of traveling salespeople [28]. To enhance the efficiency of the

algorithm and to maximize its ability to solve more complicated optimization

11
problems with more clusters or with incomplete information, Dorigo &

Gambardella improved the ant algorithm in 1997 and proposed the concept of ant

colony optimization (ACO) [29]. ACO is a probabilistic technique for finding the

optimal path. The idea is to simulate the way ants search for food in natural life.

When ants leave the colony to search for food, their paths initially wander randomly

along average lines. Ants deposit pheromones along the path trail. Since it takes

more time for ants to travel down longer paths and back again, those pheromones

have more time to evaporate. Over time, the pheromones on longer paths evaporate,

and fewer ants will follow the trail. The next time the ants choose paths, they will

choose the shorter paths which have a higher pheromone density. Since pheromone

density is the basis for identifying the shorter paths, the combined result of all

shorter paths chosen will reveal the shortest path.

Ant colony optimization is a probabilistic optimizing method. When artificial

ants optimize a path, they will consider the transition probability for the selection of

next node. The main considerations are pheromone density and length of line

segments. The formula for transition probability is shown as Eq. (1):

𝛽𝛽
[𝜏𝜏𝑖𝑖𝑖𝑖 (𝑡𝑡)]∝ ×�𝜂𝜂𝑖𝑖𝑖𝑖 �
， 𝑖𝑖𝑖𝑖 𝑗𝑗 ∈ 𝐽𝐽𝑠𝑠 (𝑖𝑖)
𝑃𝑃𝑖𝑖𝑖𝑖𝑠𝑠 (𝑡𝑡) = �∑𝑢𝑢∈𝐽𝐽𝑠𝑠 (𝑡𝑡)[𝜏𝜏𝑖𝑖𝑖𝑖 (𝑡𝑡)]∝ ×[𝜂𝜂𝑖𝑖𝑖𝑖 ]𝛽𝛽 (1)
0 ， 𝑜𝑜𝑜𝑜ℎ𝑒𝑒𝑒𝑒𝑒𝑒

Here, 𝜏𝜏𝑖𝑖𝑖𝑖 (𝑡𝑡) refers to the pheromone density of the line segment (𝑖𝑖, 𝑗𝑗) at time

𝑡𝑡. 𝐽𝐽𝑠𝑠 (𝑖𝑖) refers to the set of neighboring nodes that ant 𝑠𝑠 has not yet visited at node

𝑖𝑖. 𝜂𝜂𝑖𝑖𝑖𝑖 refers to the expected value, usually a reciprocal of the length of line segment

( 𝑖𝑖, 𝑗𝑗 ). ∝ and 𝛽𝛽 refer to the parameters determining the relative importance

between pheromone and distance; usually the value of 𝛽𝛽 is greater.

Pheromone evaporation refers to the phenomenon in which pheromones

12
deposited on the ground after the ant has chosen its path dissipate over time. In an

ant algorithm, pheromone evaporation in the natural world is emulated to avoid the

unlimited accumulation of pheromones on a certain path, which can result in

artificial ants constantly exploring other poor solutions.

Therefore, after an artificial ant selects all nodes, a solution is constructed and

the pheromone density on the paths must be updated. After all artificial ants

complete a journey, an “overall updating method” is executed to set the pheromone

density. This pheromone updating is conducted according to path performance. The

formula is shown as Eq. (2):

𝜏𝜏𝑖𝑖𝑖𝑖 (𝑡𝑡 + 1) = (1 − 𝜌𝜌)𝜏𝜏𝑖𝑖𝑖𝑖 (𝑡𝑡) + ∑𝑆𝑆=1~𝑛𝑛 ∆𝜏𝜏𝑖𝑖𝑖𝑖 𝑠𝑠 (2)

Here, 𝜏𝜏𝑖𝑖𝑖𝑖 (𝑡𝑡) refers to the pheromone density from node 𝑖𝑖 to node 𝑗𝑗 at time 𝑡𝑡.

𝜌𝜌 refers to the pheromone evaporation coefficient ( 0 < 𝜌𝜌 < 1 ). ∆𝜏𝜏𝑖𝑖𝑖𝑖 𝑠𝑠 is the

remaining pheromone density of ant 𝑠𝑠 on line segment (𝑖𝑖, 𝑗𝑗). The calculation is

shown as Eq. (3):

𝑄𝑄
𝑠𝑠
∆𝜏𝜏𝑖𝑖𝑖𝑖 = � 𝐿𝐿𝑠𝑠 (3)
0

𝑄𝑄 is the pheromone density each ant can secrete. 𝐿𝐿𝑠𝑠 is the solution obtained

by the 𝑠𝑠th ant. Each time a pheromone update is done, an iteration is deemed

complete. When the optimal solution generated no longer changes or a certain

number of iterations is reached, the ant colony optimization is complete.

Ant colony optimization has been widely used to solve various engineering

optimization problems, and some studies have applied it to tourism. Claes and

Holvoet (2011) used ant colony optimization to solve schedule planning issues and

construct a transportation system [30]. ACO can predict the best path and transit

time to estimate costs in terms of time and money. Ant colony optimization has been
13
shown to outperform the traditional static routing algorithm. Claes and Holvoet

(2012) improved the ant colony optimization by adding pheromone updating to

transportation routes for more accurate route planning and prediction, resulting in

more intelligent traffic systems [31]. Although many algorithms can solve the

shortest path problem, after evaluating the options in regards to efficiency and

accuracy, the majority of studies have adopted ant colony optimization [32, 33].

Using this algorithm for tour planning helps users conveniently and effectively find

the best path.

3. System Design and Performance Evaluation

3.1. System Platform and Architecture

The personalized location-based mobile tourism application (PLMTA)

developed in this study is based on the LBS concept. The user’s geographical

location is fed to the system via the smartphone’s positioning function. After the

system applies filtering and matching, the resulting data is written to the database

and appropriate information is then presented to the user. This study adopted

Notepad as the development platform, JavaScript as the main system programming

language, and HTML5 and CSS as the screen layout tools,. Phone Gap was used to

present the system in app format. This study also adopted Google Spreadsheet as the

cloud database for the system using the Google Maps API.

Using Google Spreadsheet provides the system with such cloud computing

advantages as cost-effectiveness, server cooperation, on-demand provisioning, and

geographic diversity. It is also allows us to build a more dynamic and cost-effective

information management infrastructure [34]. The PLMTA database is divided into

two parts: one is to collect information on attractions and iBike stations in the

14
Taichung area, including attraction (or station) profiles, location coordinates,

pictures, related videos and audios, or introductory information. The other is to

record users’ personal attraction preferences and the list of attractions selected as the

reference for future attraction filtering and recommendations.

3.2. System Execution Screen

The system is mainly divided into four parts: attraction and iBike station

information, theme travel, tour planning, and information on neighboring stores.

The functions are described as follows.

(1) Attraction and iBike Station Information

“Attraction Information” is designed to collect information regarding

attractions in the Taichung area and provide various kinds of travel information. The

execution screen is shown in Figs. 1-3. Taichung area attractions and iBike stations

and their related information are collected and organized to assist users to obtain

travel information more conveniently. This function is divided into three modes: list

of attractions, list of stations, and map mode. In addition to pictures and text

descriptions, the attraction information includes recommendation information

regarding neighboring attractions, and links to blogs or video and audio files.

15
Figure 1. List of Attractions Figure 2. List of Stations

Figure 3. Map Mode

16
(2) Theme Travel

“Theme Travel” is designed to provide several travel themes based Taichung’s

famous attractions and unique features. The system can recommend differently

themed, pre-planned tours, and users can pick from a variety of interesting

itineraries.

(3) Tour Planning

“Tour Planning” is designed to help users effectively plan their own itinerary.

Tour planning can be divided into three parts: user reference and information

inquiry, attraction filtering and recommendation result, and route planning. First a

questionnaire survey is conducted to gather the user’s attraction preferences, as

shown in Fig. 4. If the user has used the system before, it will automatically obtain

previously-noted preferences. After the user’s attraction preferences are confirmed,

preliminary filtering is done based on attraction, as shown in Fig. 5.

To ensure that user preferences do not overly restrict the recommendation

results, the system will also recommend different attractions chosen by user groups

with similar preferences. After the recommended attraction list is provided and the

user has selected the attractions to be visited, route planning is accomplished via the

ant colony optimization (ACO) algorithm, as shown in Fig. 6. Finally, the system

generates a recommended itinerary and displays it to the user. The electronic map

interface using the Google Maps API gives users a clear view of travel time and

distance. It also lists nearby iBike stations and other related information for user

reference, as shown in Fig. 7.

17
Figure 4. Tour Planning with user preferences Figure 5. Tour Planning with attraction filtering

Figure 6. Recommendation Results Figure 7. Detail Information for route planning

18
(4) Information about Neighboring Stores

“Neighboring Stores” is designed to provide other related information such as

nearby restaurants and hotels. This uses the Googles Places extension function of

the Google Maps API to provide information about neighboring stores so that users

can more conveniently obtain information regarding food, clothing, housing and

transportation in the local area.

3.3. System Performance Evaluation

Authoritative U.S. survey agency J.D. Power (2013) found that phone

performance is the key factor that consumers value the most in regards to

smartphones [35]. Mobile app developers always take performance into

consideration. Apps that close unexpectedly or crash the phone lower users’

evaluations of and satisfaction with mobile apps. This research chose one

optimizing algorithm for tour planning. Based on the results of various algorithms

evaluated in the previous section, ant colony optimization was found to be

remarkably efficient at solving traveling salesperson problems; thus this research

adopted ant colony optimization as the basis of the tour planning system. This

algorithm allows the overall operation of the mobile app to be more efficient while

maintaining the smoothness of the smartphone application.

Based on the literature and our own testing, we configured our system with

the most appropriate parameter settings as follows. The pheromone factor (∝) was

set to 1, the visibility factor (𝛽𝛽) was set to 2, and the pheromone evaporation rate (𝜌𝜌)

was set to 0.1 [29, 36, 37]. This research assessed the efficiency of ant colony

optimization for route planning. Five, seven and ten attractions were randomly

chosen for solving the TSP (traveling salesperson problem) 50 times, respectively.
19
The results are shown in Table 1.

Table 1. System Effectiveness Evaluation

Unit: millisecond (ms)

5 Attractions 7 Attractions 10 Attractions

Average Value 319.48 321.56 327.96

As shown in Table 1, the differences in solving efficiency for 5 attractions, 7

attractions, and 10 attractions are not significant. Continuing advances in hardware

and network equipment have increased people’s speed requirements for network

services or apps. Waiting times within 0.1 seconds give users the sense that the app

completes tasks instantaneously, providing the perfect operation speed experience.

When waiting times rise to around one second, users begin to feel like they are

waiting, but they usually are willing to do so since the wait time is within the

acceptable range. Therefore, most systems or apps set their optimization goal to 1.0

seconds. Once the wait time exceeds 5 seconds, users are no longer satisfied and

their intention to use the system again is dramatically reduced. The processing time

of the route planning function in this study, under all three scenarios, is less than

one second, which is acceptable to users.

4. Assessment of Intention to Use

This study explores users’ intention to use the PLMTA and establishes a

research model and hypotheses. Data collection for analysis is conducted via a

questionnaire. This research conducts confirmatory factor analysis (CFA) and

structural equation modeling (SEM).

4.1. Research Model

20
 Many studies have explored users’ willingness to accept technology based on

technology acceptance models. Some studies have incorporated other factors as

external variables in order to increase the explanatory power of the model and take

more potential factors into account. To design a research model to explore user’s

intention to use PLMTA, this study adopts the technology acceptance model (TAM)

as the theoretical basis and combines it with the information system success model

(ISSM). Information quality, system quality and perceived convenience are

considered as external variables.

(1) Information system success model

The information system success model was proposed by DeLone & McLean in

1992 [38]. Six factors are involved in evaluating the success of an information

system: system quality, information quality, system use, user satisfaction, the effect

on the individual user, and the effect on the organization.

Information quality refers to the accuracy, reliability, completeness, timeliness,

and correlation of the data produced by the information system [38]. According to

Seddon & Kiew (2007), information quality affects perceived usefulness, thereby

affecting users’ satisfaction with the information system [39]. Montazemi &

Qahri-Saremi (2015) found that information quality has a strong influence on

perceived usefulness, thereby increasing the intention of users to continue to adopt

online banking service [40]. When information quality is better, users find the

output information to be more helpful and are thus willing to use the information

system more frequently [41, 42].

System quality refers to the essential features of the system that produce

information. Chen & Hsiao (2012) explored user adoption of medical systems and

found that system quality has a strong influence on perceived ease of use, thereby

having a strong influence on user adoption intention [43]. Some studies have found
21
that system quality has a positive effect on perceived ease of use, thereby affecting

users’ willingness to shop online [44, 45].

(2) Perceived Convenience
Many scholars have proposed different perspectives regarding convenience.

From a marketing perspective, Brown (1989) stated that “convenience” can be seen

as a type of marketing strategy applied to consumer products [46]. Providing

consumers with products or services that give them a sense of “convenience” will

increase their intention to purchase.

Perceived convenience refers to the degree to which users feel the products or

services are easily found and used. Since this sense of convenience felt by users is

not constrained to time, convenient operation of the product or service should also

be considered. Hsu & Chang (2013) explored users’ acceptance of the Moodle

teaching platform and found that perceived convenience has a strong influence on

perceived usefulness, thereby affecting the intention to continue to use the system

[47]. Yoon & Kim (2007) evaluated users’ intention to accept a wireless LAN and

found that perceived convenience has a positive effect on perceived usefulness,

thereby affecting behavioral intention [48]. Tang & Chiang (2009) explored factors

affecting behavioral intention regarding mobile environment knowledge

management and found that perceived convenience is a major factor that positively

affects perceived ease of use, thereby affecting usage attitudes and behavioral

intention [49].

(3) Technology Acceptance Model

Evolving from the Theory of Reasoned Action (TRA), the TAM was proposed

by Davis in 1989, [50]. The TAM has been used to interpret users’ acceptance of

information technology. It comprises three dimensions: perceived ease of use,

perceived usefulness, and usage behavioral intention. First, the perceived usefulness
22
represents the degree of users believe the technology usage will enhance their job

performance. When perceived usefulness is high, users believe that the information

technology will help them perform their jobs better, thus increasing their intention

to use the information technology. Second, the perceived ease of use represents the

degree of complexity of the information technology. When the complexity of the

information technology is lower, the system is easier to operate, increasing users’

intention to use the information technology.

Many studies have shown that perceived ease of use and perceived usefulness

have a strong influence on the intention to adopt the novel technology, and that

perceived ease of use has a strong influence on perceived usefulness [51, 52]. Some

studies have included external factors along with the TAM for exploration purposes.

Egea & Gonzalez (2011) incorporated sense of trust as an external variable in the

TAM to explore the intention to use an electronic medical record system [53].

This research explores users’ intention to use the proposed PLMTA system,

combining the TAM and the ISSM, and incorporating information quality, system

quality, and perceived convenience as external variables. The research model is

shown in Figure 8:

Figure 8. Research Model

23
4.2. Research Hypotheses
This research uses the TAM as a basis, adding information quality, system

quality and perceived convenience as external variables. The following hypotheses

are proposed.

H1: The information quality of the PLMTA will positively influence on users’

intention to use the system.

H2: The information quality of the PLMTA will positively influence on users’

perceived ease of use.

H3: The system quality of the PLMTA will positively influence on users’ perceived

usefulness.

H4: The system quality of the PLMTA will positively influence on users’ perceived

ease of use.

H5: Users’ perceived convenience regarding the PLMTA will positively influence

on perceived usefulness.

H6: User’s perceived convenience regarding the PLMTA will positively influence

on perceived ease of use.

H7: Users’ perceived ease of use regarding the PLMTA will positively influence on

perceived usefulness.

H8: User’s perceived usefulness regarding the PLMTA will positively influence on

their intention to use the system.

H9: User’s perceived ease of use regarding the PLMTA will positively influence on

their intention to use the system.

24
4.3. Questionnaire Design and Descriptive Statistic
This study explores users’ intention to use the PLMTA and uses a questionnaire

approach for data collection. The questionnaire is divided into 7 dimensions and has

a total of 25 questions. All items are measured using a 5-point Likert-type scale (1 =

strongly disagree, and 5 = strongly agree). This study used the “mysurvey” platform

to create the questionnaire online and released it in popular forums. The total

number of questionnaires collected was 213. After eliminating invalid samples, 176

valid samples remained. Males accounted for approximately 62% of the respondents,

while females accounted for 38%. The majority of the respondents were 21-30 years

old (approximately 66%). The majority of the respondents had a college degree

(70%). Students accounted for 73% of the sample.

Gorsuch (1983) suggested that the sample size should be larger than 5 times

the number of questions, and greater than 100 [54]. This research meets the criteria.

The survey questions and their average value are shown in Table 2. The average

value of each question is greater than 3.5, indicating that users have good feedback

on each factor of the system.

Table 2. Survey Questions and Average Value

Survey Questions Average
Dimension
Value
IQ1: I think the information provided by this app is
3.713
accurate and credible.
Information IQ2: I think the information provided by this app is
3.828
Quality complete and informative.
(IQ) IQ3: I think this app responds to my inquired information
3.655
quickly and instantly.
IQ4: Overall, I am satisfied with the information quality 3.695

25
 of this app.
SQ1: I think this app really responds to my needs. 3.672
System
SQ2: I think this app allows me to operate specific
Quality 3.598
features (such as navigation) conveniently.
(SQ)
SQ3: I think the system architecture of this app is logical. 3.684
PC1: I think this app makes me saving a lot of time in
3.707
obtaining information.
PC2: I think this app makes it more convenient for me to
3.718
plan itinerary.
Perceived PC3: I think this app’s searching for information
Convenience according to my current location makes it easier to find 3.736
(PC) information as I expected.
PC4: I think this app assists me to plan itinerary more
3.701
conveniently.
PC5: I think this app makes it easier to collect
3.770
information.
PEOU1: I think I can easily operate this app. 3.851
PEOU2: I can very quickly learn how to operate this app. 3.891
Perceived PEOU3: I can very quickly operate this app to get
3.914
Ease of Use information I need.
(PEOU) PEOU4: Overall, I think this app is easy to use. 3.891
PEOU5: I think the operation interface of this app is easy
3.833
and straightforward.
PU1: I think using this app can help me obtain
3.753
information I need.
PU2: I think using this app can increase the efficiency of
Perceived of 3.787
information collection.
Usefulness
PU3: I think how app incorporated iBike information
(PU) 3.747
helps me in tour planning.
PU4: Overall, I think using this app is useful.
3.764

IU1: I will give priority to using this app for information

3.466
collection.
Intention to IU2: I think using this app for collecting information is
3.483
Use the right choice.
(IU) IU3: I will increase the frequency with which I use this
3.431
app.
IU4: I will continue to use this app in the future. 3.425
26
 5. Result and Analysis

5.1. Reliability and Validity Analysis

For reliability analysis, a Cronbach’s Alpha of greater than 0.7 indicates the

homogeneity of questions in the same dimension [55]. The reliability analysis of

this research is shown in Table 3, and all reliability values are greater than 0.7,

showing a high degree of reliability and internal consistency.

This research assessed the convergent validity and discriminant validity of the

constructs. Table 3 shows that the factor loading of each dimension is greater than

0.5, thus confirming convergent validity. In this study, although factor loadings for

each dimension do not all reach the level of “excellent” (i.e., 0.71), they can be

considered “good” (>0.63), as shown in Table 3. Bagozzi & Yi (1988) [56], and

Comrey & Lee (1992) [57] all suggested that the average variance extracted (AVE)

of potential variables should preferably be greater than 0.50, and component

reliability should be greater than 0.7. For discriminant validity to be ensured, the

degree of correlation between dimensions must be smaller than that within each

dimension. Therefore, a Pearson correlation coefficient matrix is used for validation.

In Table 4, the square root of AVE of potential variables is greater than correlation

coefficient of other dimensions, ensuring good discriminant validity [58].

Table 3. Reliability and Validity Analysis

Dimensions and Questions Factor Component AVE Value Cronbach’s
Loading Reliability Alpha
IQ1 0.814
IQ2 0.708
Information Quality 0.8323 55.47% 0.880
IQ3 0.758
IQ4 0.693
SQ1 0.735
System Quality SQ2 0.744 0.7959 56.53% 0.812
SQ3 0.776
Perceived PC1 0.718 0.8357 50.44% 0.897
27
Dimensions and Questions Factor Component AVE Value Cronbach’s
Loading Reliability Alpha
Convenience PC2 0.720
PC3 0.709
PC4 0.679
PC5 0.724
PEOU1 0.741
PEOU2 0.726
Perceived Ease of
PEOU3 0.760 0.8498 53.10% 0.907
Use
PEOU4 0.703
PEOU5 0.712
PU1 0.851
Perceived PU2 0.805
0.8884 66.57% 0.891
Usefulness PU3 0.780
PU4 0.826
IU1 0.692
IU2 0.767
Intention to Use 0.8140 52.28% 0.920
IU3 0.722
IU4 0.709

Table 4. Discriminant Validity

Information System Perceived Perceived Perceived Intention
Quality Quality Convenience Ease of Use Usefulness to Use
Information
0.745
Quality
System
0.346 0.752
Quality
Perceived
0.656 0.398 0.710
Convenience
Perceived
0.655 0.387 0.699 0.729
Ease of Use
Perceived
0.605 0.366 0.664 0.678 0.816
Usefulness
Intention to
0.641 0.448 0.704 0.728 0.75 0.723
Use
※ The diagonal is the square root of AVE within the dimension.

28
5.3. Goodness-of-Fit of Model and Structural Modeling Analysis
Before SEM analysis, the goodness-of-fit of the model must first be evaluated

to determine whether the test model and hypotheses are suitable. In this study, χ2

(176) is 449.964, CMIN/DF is 1.692 and the goodness-of-fit index (GFI) is 0.844.

The adjusted goodness-of-fit index (AGFI) is 0.810, and the incremental fit index

(IFI) is 0.940. The comparative fit index (CFI) is 0.941 and the root mean square

error of approximation (RMSEA) is 0.063. Hair et al. (1988) [58] suggested that

GFI should be greater than 0.90 and AGFI should be greater than 0.80. Nevertheless,

some studies have pointed out that the fit is acceptable as long as both GFI and

AGFI are greater than 0.80. An RMSEA value between 0.05 and 0.08 also indicates

an acceptable goodness-of-fit [56, 57, 58]. In this study, each indicator reaches the

suggested value; therefore this model is suitable for SEM analysis.

The results of path analysis are shown in Figure 9. The explanatory power of

perceived usefulness is 50.5%. The explanatory power of perceived ease of use is

54.2%. The explanatory power of intention to use is 60.7%. Both system quality

(H3, β=.68, p<0.001) and perceived convenience (H5, β=.28, p<0.01) have a

positively influence on perceived usefulness. However, the hypothesis regarding the

effect of perceived ease of use on perceived usefulness is not supported (H7, β=-.08,

p>0.05). Information quality (H2, β=.42, p<0.001), system quality (H4, β=.16,

p<0.05), and perceived convenience (H6, β=.60, p<0.001) all have a positively

influence on perceived ease of use. Information quality (H1, β=.23, p<0.001),

perceived usefulness (H8, β=.18, p<0.01), and perceived ease of use (H9, β=.60,

p<0.001) all have a positively influence on intention usage.

29
 Figure 9. Structural Equation Modeling

6. Conclusion

The personalized location-based mobile tourism application (PLMTA)

proposed in this study used ant colony optimization to achieve its tour planning

function, allowing the operation of the system to be more effective. The analysis

results show that tourism information based on LBS is more convenient to use.

Incorporating iBike information as auxiliary data can more effectively help users

plan their trips. The explanatory power of each dimension of the research model

proposed in this study is higher than 50% (perceived usefulness: 50.5%; perceived

ease of use: 54.2%; intention to use: 60.7%). This indicates that the research model

is applicable to explore this issue and can be a reference for follow-up studies.

System quality and perceived convenience have a positively influence on

perceived usefulness, indicating when users feel that overall operation quality is

good and they can conveniently and smoothly operate various functions of the

system, they consider the system to be more useful. System quality has the greatest

effect on users’ perception of system functionality. This implies that a greater focus

on system quality is needed for future system development. However, our

30
hypothesis regarding the positive effect of perceived ease of use on perceived

usefulness is not supported by the data. This is probably because users believe that

the system’s ease of use is not relevant to the system’s functionality. Information

quality, system quality, and perceived convenience all have a positively influence on

perceived ease of use. Three factors were found to give users the sense that the

system was easy to use, allowing them to operate the system quickly: a) the quality

(i.e., accuracy and timeliness) of the information produced by the system, b) the

stability and clarity of the system architecture, and c) the users’ perception of the

convenience brought by the system. Among these, perceived convenience has the

greatest effect. When the system design is more detailed, users tend to find the

system more convenient, enabling them to operate it more easily. The biggest factor

for usage intention to adopt the system is perceived ease of use. Thus, we infer an

easy-to-use system design allows users to master the system quickly, greatly

increasing their acceptance of the system. Furthermore, information quality and

perceived usefulness both own the positively influence on intention to use. This

means that the quality of the information and the ability of the system to help users

complete their tour planning will positively affect users’ willingness to use the

system.

The scope of the personalized location-based mobile tourism application

(PLMTA) proposed in this study is limited the Taichung area, so the amount of

tourism information is also limited. System information will be richer if the scope

can be widened to include other areas. Transportation information included in this

research is limited to iBike. The operability of the system will be better if more

public transportation information is included. In terms of data collection, since it is

difficult to control the kinds of subjects sampled via an online questionnaire, the

31
majority of our respondents were young (aged 20-30) and most were students. The

prediction results may be more accurate with a more even sample distribution. Since

this research proposes only six factors, future studies that incorporate more factors

will enhance the explanatory power of our research model.

ACKNOWLEDGMENT

The authors wish to thank the Ministry of Science and Technology (National

Scientific Council) of the Republic of China for financially supporting this research

under Contract Grants No. MOST 105-2511-S-005-001-MY3 and

NSC103-2410-H-005-031.

References

[1]. eMarketer(2014). Smartphone Users Worldwide Will Total 1.75 Billion in

[Link] March 20, 2014, from [Link]

[2]. J. Borras, A. Moreno, A. Valls, Intelligent tourism recommender systems: A

survey. Expert Systems with Applications 41 (2014) 7370-7389.

[3]. T. Batista, F. Freire, C.M. Silva, Vehicle environmental rating methodologies:

Overview and application to light-duty vehicles. Renewable and Sustainable
Energy Reviews 45 (2015) 192–206.

[4]. W.T. Chung, User Satisfaction with Automated Unmanned Rental

Systems-Using Taipei City YouBike as an Example. Department of
Information Management of Chung Hua University, unpublished Thesis, 2010.

[5]. N. Good, J.B. Schafer, J. Konstan, A. Borchers, B. Sarwar, J. Herlocker, J.

Riedl, Combining collaborative filtering with personal agents for better
recommendations. AAAI '99/IAAI '99 Proceedings of the sixteenth national
conference on Artificial intelligence and the eleventh Innovative applications
of artificial intelligence conference innovative applications of artificial
intelligence (1999) 439-446.

[6]. P. Resnick, H.R. Varian, Recommender system. Communications of the ACM

40(3) (1997) 56-58.
32
[7]. S. Loh, F. Lorenzi, R. Saldana, D. Licthnow, A tourism recommendation
system based on collaboration and text analysis. Information Technology &
Tourism 5 (2003) 157-165.

[8]. F. Ricci, H. Werthner, Case-based querying for travel planning

recommendation. Information Technology and Tourism 4(3–4) (2002)
215-226.

[9]. M. Wallace, I. Maglogiannis, K. Karpouzis, G. Kormentzas, S. Kollias,

Intelligent one-stop-shop travel recommendations using an adaptive neural
network and clustering of history. Information Technology & Tourism 6 (2003)
181-193.

[10]. Y. Huang, L. Bian, A Bayesian network and analytic hierarchy process based
personalized recommendations for tourist attractions over the Internet. Expert
Systems with Applications 36(1) (2009) 933-943.

[11]. N. Manouselis, C. Costopoulou, Analysis and Classification of Multi-Criteria

Recommender Systems. World Wide Web: Internet and Web Information
Systems 10(4) (2007) 415-441.

[12]. M. Montaner, B. López, J.D.L. Rosa, A taxonomy of recommender agents on

the internet. Artificial intelligence review 19(4) (2003) 285-330.

[13]. B.M. Sarwar, G. Karypis, J. Konstan, J. Riedl, Item-based Collaborative

Filtering Recommendation Algorithms.WWW10, May 1-5, Hong Kong, 2001.

[14]. M. Batet, A. Moreno, D. Sanchez, D. Isern, A. Valls, Turist@: Agent-based

personalised recommendation of tourist activities. Expert Systems with
Applications 39(8) (2012) 7319-7329.

[15]. W.S. Yang, S.Y. Hwang, ITravel ： A recommender system in mobile

peer-to-peer environment. Journal of Systems and Software 86(1) (2013)
12-20.

[16]. X. Zhou, S. Wu, G. Chen, L. Shou, kNN processing with co-space distance in
SoLoMo systems. Expert Systems with Applications 41(16) (2014) 6967-6982.

[17]. H.S. Chiang, T.C. Huang, User-adapted travel planning system for
personalized schedule Recommendation. Information Fusion 21 (2015) 3-17.

[18]. J. He, H. Liu, H. Xiong, SocoTraveler: Travel-package recommendations

leveraging social influence of different relationship types. Information &
Management (2016) [Link]
33
[19]. I. Cenamor, T. de la Rosa, S. Núñez, D. Borrajo, Planning for tourism routes
using social networks. Expert Systems with Applications 69(1) (2017) 1-9.

[20]. K. Virrantaus, J. Markkula, A. Garmash, Y.V. Terziyan, J. Veijalainen, A.

Katanosov, H. Tirri, Developing GIS-Supported Location-Based Services. In:
Proc. of WGIS’2001-First International Workshop on Web Geographical
Information Systems, Kyoto, Japan, (2001) 423-432.

[21]. Open Geospatial Consortium (2015). and Retrieved May 3, 2015 from the
World Wide Web：[Link]

[22]. J. Samsioe, A. Samsioe, Introduction to Location Based Services: Markets and

Technologies.” InReichwald R. (Ed), Mobile Kommunikation: Wertschopfung,
Technologien,neueDienste. GablerWiesbaden, Germany, (2002) 417-438.

[23]. A.Y.L. Chong, Mobile Commerce Usage Activities: The Role of Demographic
and Motivation Variables. Technological Forecasting & Social Change 80(7)
(2013) 1350-1359.

[24]. F. Yang, Z.M. Wang, A Mobile Location-based Information Recommendation

System Based on GPS and WEB2.0 Services, WSEAS transactions on
computers 4(8) (2009) 725-734.

[25]. Y. Yu, J. Kim, K. Shin, G.S. Jo, Recommendation system using location-based
ontology on wireless internet： An example of collective intelligence by using
‘mashup’ applications. Expert Systems with Applications 36(9) (2009)
11675-11681.

[26]. W. Husain, L.Y. Dih, A Framework of a Personalized Location-based Traveler

Recommendation System in Mobile Application. International Journal of
Multimedia and Ubiquitous Engineering 7(3) (2012) 11-18.

[27]. D. Gavalas, C. Konstantopoulos, K. Mastakas, G. Pantziou, Mobile

recommender systems in tourism. Journal of Network and Computer
Applications 39 (2014) 319-333.

[28]. M. Dorigo, V. Maniezzo, A. Colorni, Ant System: Optimization by a colony of

cooperating agents. IEEE Transactions on Systems, Man, and Cybernetics-Part
B, 26(1) (1996) 29-41.

[29]. M. Dorigo, L.M. Gambardella, Ant Colony System: A Cooperative Learning

Approach to the Traveling Salesman Problem. IEEE Transactions on
Evolutionary Computation, 1(1) (1997) 53-66.

34
[30]. R. Claes, T. Holvoet, Ant colony optimization applied to route planning using
link travel time predictions. in IEEE International Symposium on Parallel and
Distributed Processing Workshops and Phd Forum (IPDPSW), IEEE, (2011)
358-365.

[31]. R. Claes, T. Holvoet, Cooperative ant colony optimization in traffic route

calculations. in Advances on Practical Applications of Agents and Multi-Agent
Systems. 155 (2012) 23-34.

[32]. C. Chira, J. Sedano, J.R. Villar, M. Cámara, E. Corchado, Urban bicycles

renting systems: Modelling and optimization using nature-inspired search
methods. Neurocomputing 135 (2014) 98-106.

[33]. C.Y. Tsai, H.T. Chang, R.J. Kuo, An ant colony based optimization for RFID
reader deployment in theme parks under service level consideration. Tourism
Management 58 (2017) 1-14.

[34]. J.D. Power 〈Customer Satisfaction with Feature-Rich Smartphones Increases

，J.D. Power, and Retrieved
as the Segment's Popularity Continues to Rise〉
May 20, 2015 from the World Wide Web:
[Link]
on-research-volume-1-and-2013-us-wireless

[35]. J. Hoffman, Cloud Computing: An Introduction to SQL Azure, TechNet

Magazine, and Retrieved May 15, 2015 from the World Wide Web:
[Link]

[36]. T. Stutzle, H.H, Hoos MAX–MIN Ant System. Future Generation Computer
Systems 16 (2000) 889–914.

[37]. E.G. Kalayci, T.E. Kalayci, D. Birant, An ant colony optimisation approach for
optimising SPARQL queries by reordering triple patterns. Information Systems
50 (2015) 51-68.

[38]. W.H. DeLone, E.R. McLean, Information Systems Success: The Quest for the
Dependent Variable. Information Systems Research 3(1) (1992) 60-95.

[39]. P.B. Seddon, M.Y. Kiew, A Partial Test and Development of Delone and
Mclean’s Model of IS Success. Australasian Journal of Information Systems
4(1) (2007) 90-109.

[40]. A.R. Montazemi, H. Qahri-Saremi, Factors affecting adoption of online

banking: A meta-analytic structural equation modeling research. Information &

35
 Management 52(2) (2015) 210-226.

[41]. F.Y. Pai, K.I. Huang, Applying the Technology Acceptance Model to the
introduction of healthcare information systems. Technological Forecasting &
Social Change 78(4) (2011) 650-660.

[42]. K.H.A. Saeed, A.H. Sue, Examining the Effects of Information System
Characteristics and Perceived Usefulness on Post Adoption Usage of
Information Systems, Information & Management 45(6) (2008) 376-386.

[43]. R.F. Chen, J.L. Hsiao, An investigation on physicians’ acceptance of hospital

information systems: A case research International Journal of Medical
Informatics 81 (2012) 810-820.

[44]. T. Ahn, S. Ryu, I. Han, I. The impact of Web quality and playfulness on user
acceptance of online retailing. Information & Management 44 (2007) 263-275.

[45]. C.H. Liao, C.W. Tsou, User acceptance of computer-mediated communication:

The SkypeOut case. Expert Systems with Applications 36 (2009) 4595-4603.

[46]. L.G. Brown, The Strategic and Tactical Implications of Convenience in

Consumer Product Marketing. Journal of Consumer Marketing 6 (1989) 13-19.

[47]. H.H, Hsu, Y.Y. Chang, Extended TAM Model: Impacts of Convenience on
Acceptance and Use of Moodle. US-China Education Review A 3(4) (2013)
211-218.

[48]. C. Yoon, S. Kim, Convenience and TAM in a ubiquitous computing

environment: The case of wireless LAN. Electronic Commerce Research and
Applications 6 (2007) 102-112.

[49]. J.T. Tang, C. Chiang, Perceived innovativeness, perceived convenience and

TAM: Effects on mobile knowledge management. The 5th International
Workshop on Mobile Commerce and Services (2009 WMCS) (2009) 413-420.

[50]. F.D. Davis, Perceived Usefulness, Perceived Ease of Use, and User Acceptance
of Information Technology. MIS Quarterly 13(3) (1989) 319-340.

[51]. J.S. Lim, A.A. Abdulrahman, J.H. Heinrichs, K.S. Lim, Testing Alternative
Models of Individuals’ Social Media Involvement and Satisfaction. Computers
in Human Behavior 29(6) (2013) 2816-2828.

[52]. L.G. Wallace, S.D. Sheetz, The Adoption of Software Measures: A Technology
Acceptance Model (TAM) Perspective. Information& Management 51(2)

36
 (2014) 249-259.

[53]. J.M.O. Egea, M.V.R. Gonzalez, Explaining Physicians’ Acceptance of EHCR

Systems: An Extension of TAM with Trust and Risk Factors. Computers in
Human Behavior 27(1) (2011) 319-332.

[54]. R.L. Gorsuch, Factor Analysis (2nd Ed.). Hillsdale, NJ: Erlbaum, 1983.

[55]. J.C. Nunnally, Assessment of Reliability. In: Psychometric Theory (2nd ed.).
New York: McGraw-Hill, 1978.

[56]. R. Bagozzi, Y. Yi, On the evaluation of structural equation models. Journal of

the Academy of Marketing Science 16(1) (1988) 74-94.

[57]. A.L. Comrey, H.B. Lee, A first course in factor analysis. Hillsdale, NJ:
Erlbaum, 1992.

[58]. J.F. Hair, R.E. Anderson, R.L. Tatham, W.C. Black, Multivariate Data Analysis,
1998, 5th. edition.

37
 Chia-Chen Chen, Ph.D

Chia-Chen Chen is an Associate Professor of Management Information Systems at

National Chung Hsing University, Taiwan. Her current research interests include

RFID, context awareness, wireless and sensor network, e-learning, and smart living.

Dr. Chen’s research is published or is forthcoming in Information Sciences,

Information Fusion, Computers and Education, Journal of Educational Technology &

Society, The Electronic Library, International Journal of Mobile Communications,

Personal and Ubiquitous Computing, Expert Systems with Applications, International

Journal of Information Technology and Management, and a number of national and

international conference proceedings.

Jia-Lun Tsai, Mr.

Mr. Tsai is a graduate student of Department of Management Information Systems at

National Chung Hsing University, Taiwan. His current research interests include

e-learning, m-learning, and u-learning.

Chia
a‐Chen Chen
n, Ph.D

Jia‐LLun Tsai, M r.
 it built a Personalized Location-based Mobile Tourism Application (PLMTA)
 it allowed users to effectively search travel information and arrange their trip
 it showed that perceived ease of use significantly affected intention to use
PLMTA