Cloud Adoption Readiness Assessment Framework for Small and

Medium Enterprises in Developing Economies:

Evidential Reasoning Approach

Mesfin Workineh1, Nuno M. Garcia2, and Dida Midekso1

1
IT Doctorial Program Software Engineering track, Addis Ababa University, Addis Ababa, Ethiopia
2
Lusophone University of Humanities and Technologies, Lisbon, Portugal
[email protected], [email protected], [email protected]

Keywords: Cloud Computing, Adoption Readiness, Evidential Reasoning, Multiple Criteria Decision Analysis,
Uncertainty, Small and Medium Enterprises, Organizational Capabilities, Developing Economies.

Abstract: The aim of this paper is to develop Cloud computing (CC) adoption readiness assessment framework for
small and medium enterprises (SMEs) in developing economies. The benefits obtained from CC let the
SMEs in developing economies to consider CC as an alternate technological solution. These SMEs require
adoption readiness assessment framework in order to eliminate complexities during adoption and to ensure
the likelihood of adoption success. Most of the existing frameworks involve technological characteristics to
assess adoption readiness and also do not handle uncertainties of decision makers. But, technological
characteristics are not foremost indicates of adoption readiness. Therefore, this study proposes Cloud
adoption readiness assessment framework from an organizational resources perspective based on evidential
reasoning (ER) approach. The finding of this study contributes to the existing CC literature that does not
sufficiently address the degree to which an organization is ready to deploy CC from an organizational
resources perspective. It also helps the practitioner to make an informed adoption decision. Lastly, the
applicability of proposed framework is shown using case study and the result shows that the framework is
effective to assess Cloud adoption readiness of an organization.

1 INTRODUCTION among stakeholders and lack of inadequate business

and technical insight (Garrison et al., 2012). To
coordinate adoption process they must perform
Cloud computing (CC) is a model which provides
internal Cloud readiness assessment before adoption
computing resources as a utility over the Internet. (El-Gazzar, 2014). If an organization is not attained
This technology shifts the trend of owning computing the desired level of readiness before adoption, the
resources as a product to getting computing resources adoption of CC result into failure (Akande and Belle,
as a service on demand. 2014). Therefore, it is necessary to measure the
The benefits obtained from CC (Armbrust et al., degree of readiness of the organization in advance
2009; Schubert et al., 2010; Zhang et al., 2010; (Surya and Surendro, 2014).
Espadanal and Oliveira, 2012) and the paradigm shift Since resources are a key driver and barrier of
let the small and medium enterprises (SMEs) to technology adoption in developing countries (Molla
adopt CC as an alternate technological solution. and Licker, 2005), it is important to consider the
Cloud computing is an ideal technological solution adoption readiness and likelihood of adoption
for such SMEs with limited capital and human success of an organization from a resources
resource (Surendro and Fardani, 2012). perspective. But, there is a paucity of organization
The potential benefits obtained from adopting CC capabilities based Cloud adoption readiness
outweigh the risks for SMEs (Low et al., 2011; assessment models which guide SMEs in the
Madisha and Van Belle, 2011). But SMEs in developing countries. Therefore, for SMEs to adopt
developing countries are lagging behind to adopt this and benefit from CC they need clear resource-based
technology (Amponsah et al., 2016; Yeboah- assessment model (Loebbecke et al., 2011).
Boateng and Essandoh, 2014). This is because of the Most of the existing Cloud adoption readiness
obstacle they might face like uncoordinated adoption assessment models are based on technological
characteristics like ease-of-use, perceived usefulness cloud adoption readiness quantitatively from
and so on (Akande and Belle, 2014; Carcary et al., organization capabilities perspective. Hence, Multi-
2014; Idris et al., 2014; Kauffman et al., 2014). criteria decision making method which is considered
Technological characteristics are not real indicators as a better approach to avoid biases of the decision
of preparedness of an organization in the least maker is required to assess adoption readiness of CC.
developing nations (Workineh et al., 2017). The This study is in a position to explore the applicability
foremost indicators of adoption of innovation are of such an approach in the context of Cloud readiness
organizational capabilities (Azadegan and Teich, assessment.
2010; Iacovou et al., 1995). These studies also do
not handle decision makers (DMs) uncertainty
during measurement.
A cloud computing adoption readiness decision 3 CONCEPT OF ER APPROACH
involves many qualitative and quantitative criteria.
Evidential reasoning approach appears to be When DMs evaluate an object against a criterion, it
appropriate multi-criteria decision-making approach may not always be practical for them to give a certain
for handling DMs uncertainty (XU, 2012). That's and precise assessment. The ER approach employs
why this study proposed Cloud adoption readiness belief structures to elicit a decision maker’s
assessment framework based on ER approach. preferences and to handle uncertainties involved
This paper is organized as follows: in section 2, during measurement (XU, 2012).
related work is reviewed. A basic concept of Suppose A is an object to be assessed, with L
evidential reasoning approach is presented in section criteria C= {C1, C2, …, Cl, ..., CL} and N evaluation
3. The ER approach for Cloud adoption readiness grade H = { H1, H2, H3, Hn,..., HN} where Hn is the
evaluation is elaborated in section 4. Section 5 shows nth evaluation grade. Without the loss of generality,
the case study. Section 6 discusses managerial and it is assumed that Hn+1 is preferred to Hn. Let the
theoretical implication and forward future research weights of criteria are given as ω = {ω1 ω2 ,...,ωl
directions. The last section gives concluding remarks. ,...,ωL} where ωl > 0 for l=1, 2,..., L and
. Then the assessment of alternative A on criterion
Cl, which is denoted as S(Cl(A)), can be represented
2 RELATED WORK as S(Cl(A))={(H1, 1, l(A)), (H2, 2, l(A)), … (Hn,
n, l(A))} where n, l(A) denote the belief degree of

The trend for the adoption of CC is increasing alternative A being assessed to grade Hn on criterion
significantly from time to time and gained enormous Cl and 0 ≤ n, l(A) ≤ 1, n=1,2,…, N and
interest in research (Loebbecke et al., 2011). As a . The degree of belief refers to the
result of this several empirical studies proposed in the degree of confidence that assessed object has
literature to assess adoption readiness of an anticipated measurement grade on a particular
organization (Akande and Belle, 2014; Carcary et criterion.
al., 2014; Kauffman et al., 2014; Surya and The assessments of the K alternatives on the L
Surendro, 2014). These studies assess qualitatively criteria can be represented using belief decision
Cloud adoption readiness of an organization taking matrix (Table 1) with S(Ci( )) as its element at the
technological characteristics into consideration. kth row and ith column, where S(Ci( )) is given as:
There are very few exceptional studies which S(Ci(Ak)) = {( Hn, n, i( )), n=1, 2,…, N,
assess Cloud adoption readiness quantitatively from i=1,2,…, L, k=1,2,…, K}. The belief degree that has
organizational capabilities perspective only not been assigned to any specific evaluation grade
(Surendro and Fardani, 2012; Workineh et al., ( ), can be represented as .
2017). But these studies do not show the extent of
readiness of an organization quantitatively. Table 1: belief decision matrix.
Loebbecke et al. (2012) proposed a method for
assessing the cloud readiness of an organization’s. C1 C2 … CL
The method relies purely upon yes/no criteria and the A1 S(C1(A1)) … S(C1(A1))
decision maker’s judgment. This is results in A2
subjectivity and susceptibility to the biases or …
uncertainty of the decision makers. Ak …
The above studies do not handle decision maker
uncertainty and cannot indicate the extent of
readiness of an organization. They do not also assess
 4 ER APPROACH APPLICATION decision criterion first, the evaluation result on these
main criteria need to be transformed to decision
The ER approach consists of five major steps (Xu criterion measurement grades based on rule or utility
and Yang, 2005). In this section, these ER steps to depending on decision maker’s preferences. A rule-
assess Cloud adoption readiness of an organization based transformation is usually used to transform
were illustrated. verbal grades to a different number of verbal grades.
The quantitative criterion needs to be combined
4.1 Identification of hierarchically with other qualitative criteria in the same level so
assessment model and index that a single aggregated evaluation index generated
for the decision criterion. For instance, the capability
The criteria identified from literature (Workineh to pay for the operational expenditure of an
et al., 2017) are structured hierarchically as shown in organization needs to be transformed into five verbal
figure 1. The relative weight of these criteria and measurement grades of decision criteria. To
sub-criteria is computed using Analytic Hierarchy transform a range of values of criteria need to be
Process (AHP) method (Saaty, 1990) by decision defined as shown in table 3. The transformation
maker in table 4. Five evaluation grades were set to process needs to be done without any data loss
assess each qualitative criterion as shown in table 2. (Yang, 2001).

Table 3: Transformation of Lowest Level Criteria

Assessments to Upper Levels.
Sub-criteria/ assessment NR SR R MR CR
grades
Operational expenditure 1 2 3 4 5
in million
implementation 2 4 6 8 10
expenditure in million

Let hn+1 and hn be the values of upper and lower

evaluation grade respectively, then the distributed
degree of belief for certain quantitative input data (h)
with respect of upper and lower evaluation grade is
given as

Where are the degree of belief

associated with respect to upper and lower level
Figure 1 Cloud adoption readiness Evaluation criteria evaluation grade respectively.

Table 2: Linguistic variables and value for Evaluation 4.3 Compute basic probability mass
grade
Evaluation Grade H1 H2 H3 H4 H5
Decision makers directly assess a given alternative
Linguistic variable NR SR R MR CR
against each lower level criterion and assign belief
evaluation % < 60 60-70 70 -80 80-90 90-100
degrees to each assessment grade to measure the
grade value average 30 65 performance as shown in table 4. Then basic
75 85 95 probability mass mn,i, which represents the degree to
Where NR=Not ready, SR=Slightly Ready, R=ready,
which the ith criterion is assessed to the nth
MR=more likely ready, CR=certainly ready
evaluation grade Hn, of each lower level criterion
computed as (Xu et al., 2006; Xu and Yang, 2005;
4.2 Apply information transformation Yang and Xu, 2002a, 2002b; Yang, 2001) :
For ER algorithm to aggregate evaluation index, all
lower level criteria need to be transformed to Let mH,i be a remaining basic probability mass
associated upper level criteria measurement grade. unassigned to any individual grade for ith criterion.
To get an aggregated evaluation index for the Then mH,i computed (Xu et al., 2006; Xu and Yang,
2005; Yang and Xu, 2002a, 2002b; Yang, 2001):
 Let n=1, 2, ... , N,
, and . Then the
combined probability mass of government readiness
using the two sub-criteria is given as follow:
The unassigned basic probability mass may be )
caused due to: weights of the ith criterion ( or
th =
incompleteness of evaluation on i criterion ( ,)
(Xu et al., 2006; Xu and Yang, 2005; Yang and Xu,
Where normalization factor which is used to
2002a, 2002b; Yang, 2001).
resolve the conflict and can be calculated as:
Where: = and =
For instance, government readiness has two sub-
criteria national infrastructure and regulation and
policy. Hence before computing probability mass of
government readiness first, the probability mass of
the sub-criteria has to be computed. = =1.19
In table 4 the belief degree assigned for national Hence,
infrastructure by DMs, , ,
, , and the relative Similarly, the remaining degree of belief can be
importance of criterion is calculated as:
Hence, )
=0.667x0=0, = 0.498,
, and
= 0.134
,
but =0, and
Similarly, for Regulation and policy, , )
, , , and =0
Similarly, the remaining degree of belief that is
, not assigned to any individual grade can be
calculated as:
and ,
{H}:

Because

4.4 Aggregating assessment = + =0.263

From the final combined basic probability
The ER algorithm aggregating multiple criteria assignment the combined degree of belief for
based on belief decision matrix and the evidence government readiness criteria calculated as (Yang,
combination rule of the Dempster-Shafer theory 2001; Yang and Xu, 2002a; Xu and Yang, 2005;
(Taroun and Yang, 2011; Yang and Xu, 2002a). Taroun and Yang, 2011):
Each row of the matrix represents basic probability {Hn: ,
mass related to a lower level criterion. For instance,
the basic probability mass ( ) assigned to
evaluation grade and remaining probability mass Where (Yang and Xu, 2002a)
unassigned ( ) to any individual grades for lower ,
level criteria of government readiness give belief
decision matrix (1). ,

… (1) 18,
,
 utilities of individual evaluation grades, denoted by
U(Hn), need to be estimated first. Utilities to each
Then the final distribution assessment for grade can be assigned as evenly distributed among
government readiness criteria which is generated by evaluation grade or taking the preference of decision
aggregating the two sub-criteria represented as: maker to a certain evaluation grade, the utility
S(A)={(Hn, βn (A)), n=1,2, …, N}={( not ready, function assigns a number to an evaluation or
10.65%), (slightly ready, 67.57%), (ready, 18.18%), assessment grade. For an alternative A, suppose the
(more likely ready, 0), (certainly ready, 0)}. utility of an evaluation grade Hn is U(Hn), then the
Let high level criterion has L sub-criteria which expected utility of the aggregated assessment
is assessed with five evaluation grade H= {H1, H2, is defined as follows:
H3, H4, H5}, taking basic probability mass ( )
assigned to evaluation grade and remaining
probability mass unassigned ( ) to any individual
grades after the assessments on sub-criteria, for an Note that βn denotes the lower bound of the
object to be assessed on high level criterion an likelihood that the alternative A is assessed to Hn.
assessment matrix M (2) has been developed as The upper bound of the likelihood is given by (βn
shown below. +βH ) (Yang, 2001; Yang and Xu, 2002a, 2002b).
If the assessment is incomplete or imprecise, a
utility interval can be established for distribution
… (2) assessment based on where the unassigned degree of
belief goes either to the least preferred grade
(minimum utility) or goes to the most preferred
Then, the aggregation is carried out recursively grade (maximum utility) (Yang, 2001). Suppose the
by aggregating the first row of the matrix (sub- highest preferred grade having the highest utility is
criteria) with the second row. Then this result will be n+1 and the least preferred grade having the lowest
aggregated with the third row of the matrix. This utility is n. Then the maximum, minimum, and
aggregation continues iteratively until all rows of the average expected utility of alternative A is given by:
matrix are combined in this fashion.
The more generalized version of combined
probability represented using the following equation:

Where
i= {1, 2, …, L-1}
{H}: If all the original assessments S(A) in the belief
decision matrix are complete, then βH(A)=0 and the
evaluation value of object is a point
value = = = .
From the final combined basic probability Otherwise, the value of object is an interval
assignment the combined degree of belief calculated ( ).
as follow:
{Hn}:
5 CASE STUDY
In this section, the result from using the ER
framework to assess Cloud readiness of a public
4.5 Apply utility function to transform University in Ethiopia, namely Ambo University, is
considered. Ambo University delivers its services in
After obtaining aggregated distributed belief
four campuses. Currently, it is strengthening ICT
structure, ranking or sorting alternatives or objects
office to improve its service delivery for
based on their performance may be required. But the
academician, students and other stakeholders. All
distributed belief structure is not suitable for
campuses have dedicated broadband Internet
comparing objects directly. Hence, to precisely
connection and mini data centre. According to the
evaluate the performances of the objects expected
ICT director the university has an interest to adopt Based on the assessment given by DMs the
public Cloud services for some of the services minimum, an average and maximum utility for
rendering to stakeholders to gain cost reduction and decision variable obtained from IDS as
bring agility to services’ it provides. Hence, it is 0.3422, 0.5206, and
required to assess its extent of preparedness in 0.4314. The interval
advance for successful adoption. The extent of can capture the extent of adoption readiness of an
preparedness of the university computed and expert organization. The experts in the ICT office of the
opinion on the result was obtained. University also agree with the validity of the result
To evaluate the adoption readiness of Ambo obtained.
University first, the DMs evaluate the University
against the basic criteria as shown in table 4. Then
distributed degrees of belief of assessments given by 6 DISCUSSION
the DMs are fed into a demonstration version of
intelligent decision support system (IDS),
Cloud adoption decision is a strategic decision in
implementing ER approaches (Xu et al., 2006; Xu which the decision made at the early time might
and Yang, 2005) and then aggregated results for affect the organization at the later time. For adoption
decision criteria are obtained as shown in figure 2. decisions to go well at the later time the adoption
readiness must be assessed in advance. A detailed
understanding of Cloud readiness enables an
organization to adopt cloud solutions successfully.
In order to ensure adoption readiness, SMEs in
developing nations must assess its readiness from an
organizational capabilities perspective.
The framework proposed in this study assesses
adoption readiness quantitatively from organization
resources perspective and can handle uncertainty in
DMs. It can also evaluate the extent of readiness of
Figure 2: Distributed assessment of Cloud adoption an organization more precisely and helps the DMs to
readiness for Ambo University. make an informed decision before adoption.

A quantified form of overall distributed 6.1 Theoretical and managerial

assessment is given as expected utility. The expected
implications
utility is computed based on belief degree of
evaluation grade and based on the evenly distributed This study can extend the boundary of the Cloud
utilities among evaluation grade (U(H1)=0, computing literature by (1) establishing Cloud
U(H2)=25, U(H3)=50, U(H4)=75, U(H5)=1). Hence, adoption readiness assessment framework based on
the expected utility of the assessment or the degree ER approaches, and (2) Identify organizational
of readiness of the University is computed as 0.3422. capability based hierarchically assessment model and
Since the utility for lower level assessment grade is the relative importance of each criterion. The
zero the minimum and expected utilities are equal. framework can also serve as a Practical guideline to
carry out cloud adoption readiness assessments and
to make an informed adoption decision.

6.2 Limitations and future research work

The criteria for hierarchically assessment model were
identified only from literature. For exhaustiveness of
the criteria experts from the industry have to be
involved. The intention of measuring adoption
readiness is to avoid adoption failure or to predict the
Figure 3: Extent of readiness on main criteria. likelihood of adoption success. So, in the next step
hierarchically assessment model needs to be
The extent of readiness of the University in the enhanced in order to predict the likelihood of Cloud
second level criteria is clearly shown in figure 3. adoption success of an organization.
 Table 4: The index system of Cloud Adoption readiness.
Top layer Dimension Weight Cloud capabilities/ Factors Weight Belief Degree AU
IT- 30.7% network technologies 88.9% {( 3 , 0.5), ( 4 , 0.5)}
Infrastructure enterprise systems 11.1% {( 2 , 0.3), ( 3 , 0.5)}

8.8% learning capabilities 20% {( 4 , 0.6), ( 5 , 0.4)}

organizational Top management commitments 67% {( , 0.6), ( , 0.4)}
1 2
culture and vendor management 4.1% {( 3 , 1)}
strategy
strategies 8.7% {( 2 , 0.5), ( 3 , 0.5)}
4.3% Awareness about CC 13.7% {( , 0.5), ( 3 , 0.4)}
Readiness

2
Human Knowledge and skill 62.5% {( 1 , 0.6), ( 2 , 0.4)}
Attitude 23.8% {( 2 , 0.25), ( 3, 0.75)}

Finance 23.4% payment for adopting 50% 5000000

/Economic payment for operational 50% 3000000
32.7% Government readiness 80.0%
 National infrastructure 66.7% {( 2 , 0.7), ( 3 , 0.3)}
External
 Regulation & Policy 33.3% {( 1 , 0.6), ( 2 , 0.4)}
Environment
support industries 20.0% {( 4 , 0.8)}

7 CONCLUSION
REFERENCES
For an organization to adopt successfully latest
technology like CC managers need to consider Akande, A.O., Belle, J.P.V., 2014. A Proposed
adoption readiness of an organization in advance. Framework to Assess and Increase the Cloud
Adoption readiness assessment is a key step in Computing Readiness of Financial Institutions in
feasibility analysis of technology adoption. When South Africa, in: Confluence The Next Generation
decision-maker is evaluating adoption readiness Information Technology Summit. Presented at the 5th
against a pre-determined set of criteria, he/she may International Conference of IEEE, pp. 46–52.
face some uncertainty due to lack of decision data Amponsah, R.A., Panford, J.K., Hayfron-Acquah, J.B.,
and incomplete information. But none of the methods 2016. Factors Affecting Cloud Computing Adoption in
proposed in the literature able to address the issue of
a Developing Country-Ghana: Using Extended Unified
uncertainty. Hence, a framework which can handle
Theory of Acceptance and Use Of Technology
such kind of problems and helps DMs to make an
(UTAUT2) Model. IRJET 3.
informed decision is needed. This study proposed
Armbrust, M., Fox, A., Griffith, R., Joseph, A.D., Katz,
Cloud adoption readiness assessment framework
based on ER approach to fill this gap. The proposed R., Konwinski, A., Lee, G., Patterson, D., Rabkin, A.,
framework helps DMs in identifying the extent of Stoica, I., Zaharia, M., 2009. Above the clouds: A
readiness of an organization in each criterion and in Berkeley view of cloud computing (Technical Report
addressing areas that need to be improved before UCB/EECS-2009-28 No. 4). EECS Department,
adoption more precisely. As result organization can University of California, Berkeley.
avoid unsuccessful adoption. Unlike others readiness Azadegan, A., Teich, J., 2010. Effective benchmarking of
assessment framework, which judging an innovation adoptions: A theoretical framework for e-
organization simply as ready or not ready, the one procurement technologies. Benchmarking: An
proposed in this research clearly shows the extent of International Journal 17, 472–490.
readiness of an organization in each criterion Carcary, Marian, Eileen Doherty, Gerard Conway, and
quantitatively. Therefore, the framework proposed in Stephen McLaughlin. 2014. “Cloud Computing
this study is found out as an appropriate methodology Adoption Readiness and Benefit Realization in Irish
for Cloud adoption readiness assessment and for the SMEs—An Exploratory Study.” Information Systems
decision makers to make an informed decision. Management 31 (4): 313–27.
Carroll, N., Helfert, M., Lynn, T., 2014. Towards the
Development of a Cloud Service Capability
 Assessment Framework, in: Continued Rise of the Surendro, Kridanto, and Adiska Fardani. 2012.
Cloud: Advances and Trends in Cloud Computing. “Identification of SME Readiness to Implement Cloud
Springer, London, pp. 289–336. Computing.” In International Conference on Cloud
El-Gazzar, R.F., 2014. A Literature Review on Cloud Computing and Social Networking, 1–4. Bandung,
Computing Adoption Issues in Enterprises, in: West java, Indonesia: IEEE.
Creating Value for All Through IT. Springer, Berlin Surya, G.S.F., Surendro, K., 2014. E-readiness framework
Heidelberg, pp. 214–242. for cloud computing adoption in higher education, in:
Espadanal, M., Oliveira, T., 2012. Cloud Computing International Conference of Advanced Informatics:
Adoption by firms, in: MCIS 2012. Concept, Theory and Application. IEEE, pp. 278–282.
Garrison, G., Kim, S., Wakefield, R.., 2012. Success Taroun, A., Yang, J.-B., 2011. Dempster-Shafer theory of
Factors for Deploying Cloud Computing. evidence: potential usage for decision making and
Communications of the ACM 55, 62–68. risk analysis in construction project management.
Iacovou, C.L., Benbasat, I., Dexter, A.S., 1995. Electronic Journal of the Built and Human Environment 4, 155–
data interchange and small organizations: Adoption 166.
and impact of technology. MIS Quarterly 19, 465-485. Workineh, Mesfin, Nuno M. Garcia, and Dida Midekso.
Idris, A.S., Mokhsin, M., Fauzi, F.H.M., 2014. The 2017. “Cloud Computing as Technological Solutions
Readiness Of Cloud Computing: A Case Study In for Higher Education Institutions: Adoption Readiness
Politeknik Sultan Salahuddin Abdul Aziz Shah, Shah Assessment Model: Research in-Progress.” In ICCNI,
Alam, in: International Conference on Computational 1–7. Lagos, Nigeria: IEEE.
Science and Technology – 2014. pp. 1–5. XU, D.-L., 2012. An introduction and survey of the
Kauffman, R.J., Ma, D., Yu, M., 2014. A Metrics Suite for evidential reasoning approach for multiple criteria
Firm-Level Cloud Computing Adoption Readiness, decision analysis. Ann. Oper. Res. 195, 163–187.
in: Economics of Grids, Clouds, Systems, and Xu, D.-L., McCarthy, G., Yang, J.-B., 2006. Intelligent
Services. Springer International Publishing, pp. 19–35. decision system and its application in business
C. Loebbecke, B. Thomas, and T. Ullrich. 2012. innovation self-assessment. DSSs 42, 664–673.
“Assessing Cloud Readiness at Continental AG” MIS Xu, D.-L., Yang, J.-B., 2005. Intelligent decision system
Quartely Executive, 11 (1): 11–23. based on the evidential reasoning approach and its
Loebbecke, C., Thomas, B., Ullrich, T., 2011. Assessing applications. Journal of telecom. & IT 3, 73–80.
cloud readiness: Introducing the magic matrices Yang, J.-B., 2001. Rule and utility based evidential
method used by Continental AG, in: Governance and reasoning approach for multiple attribute decision
Sustainability in Information Systems. Presented at the analysis under uncertainty. European Journal of
Managing the Transfer and Diffusion of IT, Springer, Operational Research 131, 31–61.
Berlin Heidelberg, pp. 270–281. J.-B. Yang and D.-L. Xu,. 2002b. “On Evidential
Low, C., Y. Chen, M. Wu, 2011. Understanding the Reasoning Algorithm for Multiple Attribute Decision
determinants of cloud computing adoption. Industrial Analysis under Unceratinity.” IEEE Transaction on
Management & Data Systems 111, 1006–1023. Systems, Man , and Cybernetics-Part A: System and
Madisha, M., Van Belle, J.P., 2011. Factors Influencing Humans 32 (3): 289–304.
SaaS Adoption by Small South African Organisations, Yang, J.-B., Xu, D.-L., 2002b. Nonlinear information
in: 11th Annual Conference on World Wide Web aggregation via evidential reasoning in multiattribute
Applications. Durban, South Africa. decision analysis under uncertainty. IEEE
Molla, A., Licker, P.S., 2005. eCommerce adoption in Transactions on Systems, Man and Cybernetics, Part
developing countries: a model and instrument. A: Systems and Humans, 32, 376–393.
Information & Management 42, 877–899. Yeboah-Boateng, E.O., Essandoh, K.A., 2014. Factors
Rockmann, Robert, Andy Weeger, and Heiko Gewald. Influencing the Adoption of Cloud Computing by
2014. “Identifying Organizational Capabilities for the Small and Medium Enterprises in Developing
Enterprise-Wide Usage of CC.” In PACIS 2014. Economies. IJESE 2, 13–20.
Saaty, T., 1990. How to make a decision: The Analytic Zhang, Q., Cheng, L., Boutaba, R., 2010. Cloud
Hierarchy Process. Eur. J. Oper. Res 48, 9–26. computing: state-of-the-art and research challenges.
Schubert, L., Jeffery, K., Neidecker-Lutz, B., 2010. The Journal of internet services and applications 1, 7–18.
future of cloud computing: Opportunities for European
cloud computing beyond 2010 (Expert Group report,
public version 1).