Web Migration
A Survey Considering the SME Perspective
Sebastian Heil and Martin Gaedke
Technische Universit
¨
at Chemnitz, 09107 Chemnitz, Germany
Keywords:
Web Migration, Systematic Mapping Study, SME, Software Migration, Web Engineering.
Abstract:
Legacy systems are business-critical and contain valuable knowledge gathered over years of development.
Many of these systems are still non-web desktop applications. Companies are aware of the advantages of web
applications. However, Web Migration (WM), i.e. transforming legacy systems into web applications, is still a
challenge, in particular for Small and Medium-sized Enterprises (SMEs). A large body of research exists in
this field, but the specifics of SMEs have been overseen so far. This survey provides an overview on existing
WM approaches with a dedicated focus on the SME perspective. The systematic mapping study analyses 122
primary studies and tools, identifies four major research focuses, reports on common methods and techniques
and the consideration of SMEs. We then outline resulting research issues and future research directions.
1 INTRODUCTION
User expectations, advantages of platform-
independent deployment and fading away of
support for obsolete technologies provide a rationale
for renewing legacy software as web applications.
Web browsers are becoming the standard interface
for many applications and web applications provide
a solution to platform-dependence and deployment
issues (Aversano et al., 2001). However, keeping ex-
isting valuable knowledge that is contained in legacy
systems (LS) is crucial in enterprise contexts. It is a
key asset for the company typically representing the
result of years of requirements and domain knowl-
edge elicitation. Migration bears the risk of losing
this knowledge and is therefore not easily undertaken.
These two conflicting motivations are the origin of web
migration research, which provides solutions allowing
migrating to the Web without losing knowledge.
Web migration (WM) is a challenge for Small and
Medium-sized Enterprises (SMEs) in particular, due to
their characteristics and constraints (Nussbaumer and
Liu, 2013). SMEs have limited human resources, typi-
cally less than 250 employees (European Commission,
2003). Therefore, it is not feasible to apply a migration
approach which requires an entire development team
exclusively dedicated to migration work items. SMEs
typically target a very specific market segment with a
limited customer base (Rose et al., 2016). In contrast
to large enterprises, they have less financial resources
(Nussbaumer and Liu, 2013) and no diverse portfo-
lio of products/services to provide them with several
sources of income. Thus, the risk of a failed migration
is substantially higher. SMEs are not just little big busi-
nesses
1
. Regarding human resources, collaboration is
often better but there is a lack of specialized expert
professionals (Yew Wong and Aspinwall, 2004).
The relevance of SME companies has been ac-
knowledged and expressed through various funding
programs both in the EU
2
and USA
3
. However, we
were surprised to not find much work addressing WM
for SMEs. Existing surveys (Khadka et al., 2013; Raza-
vian and Lago, 2010; Spanos et al., 2012; Jamshidi
et al., 2013; Gipp and Winter, 2007; Kienle and Dis-
tante, 2014) in this field do not sufficiently address the
characteristics of SME software companies.
Therefore, this survey aims at providing an
overview on existing WM approaches and introduc-
ing the SME perspective to this field by providing the
following contributions:
• Overview of the state of art in web migration
• Identification of the degree of considera-
tion/support of SMEs in web migration
1
https://hbr.org/1981/07/a-small-business-is-not-a-little-
big-business
2
https://ec.europa.eu/programmes/horizon2020/en/h2020-
section/sme-instrument
3
https://www.sba.gov/offices/headquarters/oca/resources
/6827
Heil, S. and Gaedke, M.
Web Migration - A Sur vey Considering the SME Perspective.
DOI: 10.5220/0006353502550262
In Proceedings of the 12th International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE 2017), pages 255-262
ISBN: 978-989-758-250-9
Copyright © 2017 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
255

Table 1: Search Terms.
Aspects Search Terms
Migration in general
Migration Migration, Transformation,
Modernization, Re(-)engineering
Migration types
UI UI, User Interface
Program Program, Code
Data Data, Database
Interface Interface, API
System System, Software, Application
Application types
Legacy Legacy
Target Web, Cloud
The remainder of this paper is organized as follows: In
section 2 we outline the research method, in section 3
we report on the results and section 4 concludes the
paper and gives an outlook on following work.
2 RESEARCH METHOD
In this systematic mapping study, we follow the guide-
lines provided by (Petersen et al., 2015).
2.1 Research Questions
To answer the initial question of “How can SMEs
feasibly conduct a migration of legacy software to the
web?”, we attempt to identify types of Web Migration
approaches and assess their applicability for SMEs,
resulting in the following research questions:
Q1 Which approaches/methods/models/processes/tools
for migration of legacy to web applications exist?
Q2 Where and when were they published?
Q3 Which aspects of WM are addressed by them?
Q4 How can they be categorized?
Q5 Which approaches explicitly consider the situation
and constraints of SMEs and how do they do so?
2.2 Data Sources & Search Strategy
In order to identify the primary studies available to
answer the research questions, we used ACM Digital
Library, IEEE Xplore, Science Direct, Scopus, Web of
Science and Wiley Online Library as data sources.
Based on existing literature, we gathered search
terms for migration in general, migration types and
application types by listing synonyms as shown in
table 1. We constructed search queries combining the
alternative search terms as in table 2. These queries
Table 2: Search Queries.
Area Query
UI (Migration OR Transformation OR
Modernization OR Reengineering)
AND (UI OR User Interface) AND
(Web OR Cloud) AND Legacy
Code (Migration OR Transformation OR
Modernization OR Reengineering)
AND (Program OR Code) AND (Web
OR Cloud) AND Legacy
Data (Migration OR Transformation OR
Modernization OR Reengineering)
AND (Data OR Database) AND (Web
OR Cloud) AND Legacy
API (Migration OR Transformation OR
Modernization OR Reengineering)
AND (API OR Interface) AND (Web
OR Cloud) AND Legacy
System (Migration OR Transformation OR
Modernization OR Reengineering)
AND (System OR Software OR
Application) AND (Web OR Cloud)
AND Legacy
were executed on the above data sources considering
titles, abstracts and keywords. To limit the time frame,
we accepted only publications from year 2000 on. This
search strategy yielded 870 initial search results.
2.3 Study Selection
Study selection ensures that results which match the
search queries but are not relevant for the research
questions are disregarded and only publications of
sufficient quality are considered. To perform study
selection, we define a list of inclusion and exclusion
criteria as presented in table 3 and table 4. I1 and
E1-E3 relate to relevance and scope, I2 was added for
consideration of quality. The inclusion and exclusion
criteria are applied to titles and abstracts of the initial
search results. If no clear decision can be made, the
full text of the publication is checked. As a result of
study selection, 92 publications have been considered
as relevant primary studies.
2.4 Migration Tool Snowballing
Q1 also includes tools for migration. However, running
the queries against scientific databases did not yield
sufficient results for migration tools. Therefore, we
applied the snowballing method (Wohlin, 2014) for the
identification of tools. It uses traversal of references
of a set of publications to identify additional ones
and proceeds in three stages: Start Set, Iteration and
ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering
256

Table 3: Inclusion Criteria.
Id Criterion
I1 Publications which contribute to the migra-
tion of user interfaces, databases, code, APIs
or of an entire system to web-based systems
Motivation: elicit approaches that provide
solutions for web migration
I2 Peer-reviewed publications which have been
published in journals, as conference papers
or workshop papers
Motivation: assure quality by considering
only work that has passed review of peers
Table 4: Exclusion Criteria.
Id Criterion
E1 Publications which do not explicitly con-
tribute to the migration to web applications
Motivation: this study aims at finding web
migration solutions, publications about other
migration topics like migration to 4GL lan-
guages or to object-oriented databases are
not considered
E2 Publications which do not explicitly provide
solution for migration scenarios
Motivation: this study aims at finding web
migration solutions, publications about the
development and maintenance of web ap-
plications or about the advantages and dis-
advantages of migration in general are not
considered
E3 Introductions, summaries, secondary studies
or publications with less than three pages
Motivation: these publications do not pro-
vide sufficient information to provide objec-
tive solutions for migration scenarios
Data Extraction. The primary studies are used as start
set. To identify software tools for WM, backward
snowballing is applied. We look for references to
existing tools instead of referenced publications. Since
software tools to not reference other tools, the iteration
phase finishes after one step. Similar to study selection,
the tools are filtered. Criteria I1 and E1 are also applied
to the tools. In addition, we specify one tool specific
exclusion criterion E4 in table 5. This snowballing
method identified 30 relevant primary tools which are
added to the primary studies set.
2.5 Assessment of Approaches
The primary studies and tools are evaluated and results
captured in a spreadsheet according to these criteria:
1. Publication/Tool Metadata
2. Migration Type
Table 5: Tool-specific Exclusion Criteria.
Id Criterion
E4 Tools, which cannot be found online and are
no longer available on the web either, neither
commercially nor for free
Motivation: the tools included in this survey
should be easily accessible for SME soft-
ware companies, tools which have been ref-
erenced in publications but are no longer
available online are therefore excluded
3. Legacy and Target System
4. Migration Disciplines
5. Consideration of SMEs
Publication/Tool Metadata: captures general infor-
mation on the evaluated approaches and tools, includ-
ing title/name of the publication/tool, if applicable
name of the presented framework, authors or com-
panies involved, year of publication, URL reference,
abstract or summary and secondary sources/reports.
Migration Type. Web migration comprises all areas
of software migration: User Interface (UI), Code, Data,
System interfaces and System. However, most publi-
cations/tools focus only on a subset of these areas. If
applied in isolation, user interface, code and data mi-
gration require a strong separation of concerns in the
source legacy system (LS). System interface migration
has become increasingly important in the context of
web services, e.g. migrating from SOAP to RESTful
web services. As concerns are often not well separated
in legacy code, system migration focuses on the joint
migration of interfaces, application logic and data.
Legacy and Target System. WM approaches provide
solutions to move from a legacy to a target system.
These systems are defined by their architecture, un-
derlying paradigms and concrete technologies. For
UI migration, technologies in terms of programming
languages/libraries of the LS and scripting/markup
languages of the target system are considered. Code
migration is described by legacy and target program-
ming language. For data migration, we specify the
database management systems. Legacy and target sys-
tem interface are described by the interface paradigm
(e.g. SOAP, REST) and libraries used. For system
migration, architectural aspects (e.g. Web Services,
Cloud, MVC) are considered.
Migration Disciplines. To describe migration activ-
ities, related work identifies several process models
with varying phases/disciplines (Khadka et al., 2013;
Razavian and Lago, 2010; Jamshidi et al., 2013; Lewis
et al., 2006). In the context of this study, we follow
the Reference Migration Process (ReMiP) (Gipp and
Winter, 2007) metamodel, which provides a generic
Web Migration - A Survey Considering the SME Perspective
257

description framework adaptable to all existing migra-
tion process models. ReMiP identifies seven migration
disciplines: Requirement Analysis, Legacy Analysis,
Target Design, Strategy Selection, Implementation,
Test, Deployment.
Explicit Consideration of SMEs. Q5 investigates the
consideration of SMEs in web migration. This crite-
rion highlights approaches which are explicitly target-
ing SMEs by addressing the characteristics described
in section 1 such as limited human resources.
Appropriateness for SMEs. Even if the majority of
approaches does not explicitly consider SMEs, differ-
ent levels of appropriateness for SMEs can be identi-
fied. To achieve this, we use the following sub-criteria:
1. Evidence
2. Industrial Relevance
3. Tool support
For this quality criterion, we employ a simple overall
rating which is formed by the sums of ratings for the
sub-criteria. Sums from 0 to 2 are rated − (not suffi-
cient), 3 to 5 ± (medium) and 6 to 7 + (good). In the
following, the partial ratings are stated in brackets.
Evidence. Available evidence of application of the
approaches in this study in practice ranges from no
evidence of application in practice (0), application in
sample scenarios (1), pilot projects (2) to successful
migration of entire systems (3).
Industrial Relevance. For evidence of application
in practice, this distinguishes between application in
industrial (1) or other, e.g. academic, contexts (0).
Tool Support. The available tool support is an impor-
tant factor for SMEs. Not all approaches offer tool
support (0). If there are tools for the given approach,
these can be either just described in publications and
not publicly available (1), or publicly available either
commercially (2) or for free (3). However, a dedicated
costs criterion could not be included since hardly any
of the investigated approaches provides sufficient infor-
mation for evaluation. Also, size and desired quality
of a migration have a higher impact on migration costs
than the chosen tool.
2.6 Threats to Validity
The survey was primarily designed by one researcher
and executed and validated by three, including the
designer. This introduces the chance of misinterpre-
tation by the executors, a threat to construct validity.
To address this, all executors were involved in major
design decisions and regular meetings were held to
assure a common understanding of the concepts and
process, detailed descriptions for the terms were kept
in a shared document.
Threats to internal validity are potential subjective
biases in the process of the survey. At the beginning of
our survey, we derived search terms from the research
questions. As WM is a wide research area, different
terminologies exist in academia and industry. There-
fore, we included synonyms as search terms. However,
our selection might not cover all relevant work and dif-
ferent synonyms might have yielded different studies.
To select primary studies, we employed inclusion
and exclusion criteria. Their specification influences
the scope of the analysis. To achieve transparency, we
provide the motivation for each of the criteria. Also,
publications and tools are inherently different, so we
included an additional tool-specific criterion.
The analysis and categorization of the primary stud-
ies and tools is prone to bias. To mitigate this, we
applied the categories established by ReMiP (Gipp
and Winter, 2007) for migration type and discipline.
The analysis was conducted by two researchers and
cross-checked by a third. To make the results of the
analysis comprehensive, we detailed the criteria. Since
“Appropriateness for SMEs” explicitly rates, we try to
reduce the subjective bias by breaking it down into
three sub-criteria with simple, easy to identify scales.
Considering external validity, i.e. generalizability
of results, we attempted to widen the scope of this
survey beyond academia by explicitly adding industrial
and commercial tools. However, since our start set
for snowballing were the primary studies, tools not
referred to in publications are not included.
3 RESULTS
For brevity, detailed results, additional figures and the
full data of the entire survey is available online
4
.
Q1. Out of 870 initial search results according to our
search strategy, 92 publications have been selected
as primary studies using the inclusion and exclusion
criteria described above. Additionally, the snowballing
identified 30 primary tools.
Q2. An overview of the distribution of the primary
studies is shown in fig. 1. Web migration has long been
a research interest and the number of publications per
year is increasing. The peak of 13 publications in 2013
correlates with increased research in cloud computing
and thus web migration in the context of migrating
to the cloud. The primary studies starting from 2004
show a focus on migration to SOA. Evolution of exist-
ing web systems can be found over the entire timespan.
We consider the distribution of primary studies
across venues with a minimum threshold of 3 publica-
4
https://vsr.informatik.tu-chemnitz.de/demos/WebMig
rationSurvey
ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering
258

1
2
3
5
6 6
5
4
5 5
8
10
7
13
9
3
0
2
4
6
8
10
12
14
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Figure 1: Distribution of primary studies over years.
tions. Conferences like WSE (8 publications), ICSM
(6), WCRE (6) or CSMR (5) from the software main-
tenance, evolution and re-engineering fields lead, fol-
lowed by Software Engineering venues like STEP(3)
and APSEC(3) and service-oriented computing con-
ferences like SERVICES (3). What surprised us is
that none of the Web or Web Engineering conferences
like WWW or ICWE appear in this list. Apparently,
the Web Engineering community does not consider
migration aspects to any significant extent so far.
3.1 Migration Type & Discipline
Coverage
Q3. Regarding the different migration types, out of
122 primary studies and tools, 20 addressed UI mi-
gration, 15 data migration, 2 interface migration, 7
code migration and 78 system migration. The high
percentage of system migration approaches was ex-
pected since WM requires paradigm changes affecting
all system aspects like towards asynchronous request-
response, client-server separation in the spatial and
technological dimension and URL-addressable UI
states and navigation. Among the discrete aspects,
UI is the most prominent, as it differs the most com-
pared to non-web systems. This is followed by the
data layer where, in the context of cloud computing,
document- and column-oriented NoSQL databases
have gained importance in addition to traditional rela-
tional databases.
Our survey identified the following coverage of
migration disciplines: 32 address requirement analy-
sis, 46 legacy analysis, 31 target design, 10 strategy
selection, 86 implementation, 13 test and 10 deploy-
ment. This distribution shows high numbers for dis-
ciplines also found in traditional (forward) Software
Engineering like requirement analysis and implemen-
tation. Legacy analysis, which is specific to software
migration, also has attracted a high interest. What
surprised us is the relatively low number addressing
the test discipline. This seems to be in contrast to
functional equivalence which is desirable in software
migration in general. The tools are mostly dedicated to
one or two migration disciplines and therefore limited
in their coverage of the entire migration process.
We could find six approaches (Lucia et al., 2008;
Khadka et al., 2011; Marchetto and Ricca, 2008;
Bernhart et al., 2012; Mohagheghi and Sæther, 2011;
Strauch et al., 2013) which provide full coverage of
all disciplines and can therefore be considered as com-
plete approaches with regard to ReMiP.
3.2 Methods, Techniques & Research
Focuses
Q4. There is a large variety of methods in use which
can be grouped, on top level, into three groups:
• Transformation approaches, that process the
legacy source code by a series of automatic trans-
formations turning it into the target system’s source
code,
• Reengineering approaches, that employ reverse en-
gineering techniques to extract information from
the legacy system and follow up with forward en-
gineering to create the target system and
• Encapsulation approaches, that wrap the un-
changed legacy system or parts of it and expose
a new interface which is then integrated with the
target system.
We illustrate these groups using the six approaches that
are complete wrt. migration discipline coverage. The
incremental strategy in (Lucia et al., 2008) belongs to
encapsulation approaches, creating wrappers for the
legacy backend, however, combined with UI reengi-
neering. (Khadka et al., 2011) combines the technical
and business perspective in serviciFi, using method
engineering. At the core, serviciFi employs concept
slicing (program slicing in combination with concept
assignment) followed by manual implementation. It
is a reengineering approach. (Marchetto and Ricca,
2008) is an incremental, semi-automatic transforma-
tion approach, because functionality is extracted and
Web Migration - A Survey Considering the SME Perspective
259

transformed into web services, using Axis2
5
to create
WSDL descriptions and stubs. (Bernhart et al., 2012)
presents another incremental reengineering approach,
with dedicated focus on parallel operation and a hard
downtime constraint. Based on manual legacy analysis,
functional requirements are elicited, rapid prototypes
created, tested with users and implemented. Focus-
ing on model-driven migration based on ADM, KDM,
SoaML etc., REMICS (Mohagheghi and Sæther, 2011)
combines reengineering with semi-automatic model-
to-model transformation. Progressing in moderniza-
tion sprints and employing meetings and artifacts from
Scrum, it is incremental and agile. The approach in
(Strauch et al., 2013) focuses on decision support to
provide guidance for migrating the database layer of
an application to the cloud. As it does not describe
any concrete methods or tools for transforming data
or encapsulating legacy databases, we classify it as a
reengineering approach.
Our survey revealed four main research foci in
web migration within the survey’s time frame:
• Migration to Web
• Migration to SOA
• Migration to Cloud
• Web Systems Evolution (WSE)
Migration to Web: is the most basic area of web mi-
gration. Research addresses the problem of bringing
legacy non-web systems into the web. This comprises
desktop TUI or GUI applications and mainframe sys-
tems. There is a large variety in what is considered the
target “Web state”. Some minimal approaches cover
only the creation of near-identical copies of legacy
user interfaces in HTML, typically adding a middle-
ware for handling HTTP-Requests and wrapping the
legacy backend, like (Karampaglis et al., 2014). Oth-
ers comprise a re-structuring, e.g. towards the MVC
architecture (Bodhuin et al., 2002). However, most
approaches fall short of creating state-of-the-art web
application results, e.g. with regard to rich interac-
tion, responsiveness, let alone touch support or easy
URL-based information sharing.
Another well researched area is migration to
service-oriented architectures (SOA) (Khadka et al.,
2013; Razavian and Lago, 2010). Here, the focus
is on identification of services in legacy applications
(Sosa et al., 2013) and architectural transformation
towards a loosely coupled orchestration of these web
services. This is mainly motivated by the reuse of
valuable business logic (Khadka et al., 2013). SOA
migration affects not only the technical but also the
business perspective, thus there is a strong focus on
5
http://ws.apache.org/axis2
business processes and business process modeling.
Also, an increased interest in model-driven methods,
e.g. based on OMG’s Architecture-Driven Modern-
ization (ADM)
6
and its Knowledge Discovery Meta-
model (KDM) can be seen. Comprehensive research
projects in the SOA migration area include REMICS
considering SOA migration in the context of cloud,
MIGRARIA
7
with its focus on model-driven web en-
gineering and SOAMIG
8
based on compiler technol-
ogy. Many approaches treat the legacy system as black
box and therefore create wrappers, e.g. (Canfora et al.,
2008) using interaction modeling by finite state au-
tomata. Regarding the target technologies, most of the
work addresses migration to SOAP-based web services
and related standards like WSDL.
Research in migration to the cloud focuses on
the challenges from the change in environment and the
SaaS paradigm. For instance, cloud platforms often
provide specific NoSQL databases (Cai et al., 2015)
and message queuing services for which LS have to
be transformed to employ. Another main issue is data
security, since migration to the cloud means involving
a third party and adding data flows which can be com-
promised. This is one of the reasons why often hybrid
architectures – partially on-premise and partially in
the cloud – are the target for migration. Here, it is cru-
cial to identify the distribution of components between
on-premise and cloud optimized with regard to factors
like communication costs (Huang et al., 2014).
Unlike the three research focuses described before,
Web Systems Evolution: addresses legacy systems
that are already web systems. Kienle et al. (Kienle
and Distante, 2014) provide an excellent overview, in
which they identify a chain of evolution from static
websites over rich internet applications (RIA), web
services/SOA, ajax-based web applications to cloud-
based and HTML5-based web applications. WSE ap-
proaches describe ways how to move between the dif-
ferent stages of this chain. The majority of recent web
migration research belongs to this category.
There is little work on migrating pre-web systems
to the latest evolution stages of this chain. Most re-
search into migration to the web was conducted in the
late 1990s and early 2000s. Target technologies like
JSP or Flash and the resulting migrated systems are
outdated by current standards. Modern HTML5-based
user interfaces and user interaction patterns are only
considered in the context of WSE.
6
http://adm.omg.org/
7
http://www.eweb.unex.es/eweb/migraria/
8
http://www.soamig.de/
ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering
260

3.3 SME Consideration
Q5: (Nussbaumer and Liu, 2013) is one of the few
approaches to explicitly address SMEs. This work fo-
cuses on cloud migration, naming security, reliability,
cost, performance, flexibility and service & support as
SME cloud provider selection requirements. However,
discipline coverage is low, and due to lack of sufficient
evidence, industrial relevance and available tool sup-
port, it is rated low in SME appropriateness. In spite
of a general interest in cloud computing, SME cloud
adoption is still low and Nussbaumer et al. identify
security concerns, the challenge of migrating exist-
ing applications and lack of understanding of busi-
ness processes as key problems. The study confirms
our own findings that SME-specific migration has not
yet been discussed in research and is only addressed
by proprietary commercial tools associated with high
costs. Major non-proprietary frameworks “lack suf-
ficient SME relevance” (Nussbaumer and Liu, 2013).
(Bodhuin et al., 2002) state that their approach is in
the context of “a research project aimed at defining
new technological solutions to be transferred to Small
and Medium Enterprises operating in the Information
and Communication Technologies”, (Mohagheghi and
Sæther, 2011; Lucia et al., 2006; Lucia et al., 2008)
report on use cases from SME partners without ex-
plicit consideration of SME characteristics in their
approaches. (Zhang et al., 2010) mentions SMEs for
motivating their work by naming SaaS advantages for
SMEs.
The tools in our survey rate generally high with
regard to appropriateness for SMEs (average 7.8). The
main reason are the high ratings for evidence, most
of them provide evidence of successful migration of
entire systems, and, evidently, for tool support.
4 CONCLUSIONS
Our survey of web migration and the SME perspective
has identified two main research issues:
1. Lack of WM research addressing SMEs
2. Modern Web applications only considered in WSE
Few researchers have addressed WM for SMEs, in
spite of their specific characteristics and relevance.
From our own collaboration experience with industry
partners, we know there are still many non-web legacy
applications which could benefit from WM. This field
provides challenging research problems and should
get more attention in research. Integration of WM pro-
cesses with agile development and maintenance activi-
ties, feasible WM methods for limited human/financial
resources and lightweight transformations provide op-
portunities for researchers to contribute to advancing
the state of the art in this field.
Most research into WM from non-Web to Web ap-
plications dates back to the early 2000s. Modern web
applications based on HTML5 and new technologies
like Web Components, Web Sockets, Web Storage,
Web Workers are only considered in the context of
WSE. In several research projects in collaboration with
partners from the eHealth and industrial inspection &
certification domains, we experienced that industry
is still struggling to bring non-web LS to the Web.
However, this problem has been neglected in recent
research. Existing approaches are not applicable due
to outdated technological foundations and lack of con-
sideration of current web technologies.
We feel that both issues are related as the impact of
the lack of SME-centered approaches is further inten-
sified by the missing motivation to migrate to the web
due to un-recent target systems of existing approaches,
raising the initial hurdle for SMEs to migrate to the
web. Therefore, new web migration approaches specif-
ically targeted at SMEs are required, which can be
applied with limited resources, integrated into day-to-
day development and which help to produce modern
HTML5-based web applications from non-web LS.
We plan to elaborate on an approach of this kind and
to create supporting infrastructure in future work. This
includes a more detailed specification of SME require-
ments and their impact on WM.
Web migration is an important research topic
which is seeing increasing interest. We identified four
major research focuses over the last 16 years. WM
is a widely overseen topic in the Web Engineering
community. Software Engineering and Maintenance
communities have shifted focus towards SOA, Cloud
and WSE. Legacy non-web to web approaches are
scarce. Our survey of 122 studies and tools indicates
a lack of consideration of SME characteristics. Based
on the identified research issues, we briefly outlined
potential research directions, in particular further speci-
fying SME requirements for WM and creating suitable
approaches. These should be developed and evaluated
in close collaboration with SME software providers of
different domains and origins.
ACKNOWLEDGEMENTS
The authors would like to thank Annika Krebs and
Sven Maak for their valuable contributions to con-
ducting this survey. This research was supported by
the eHealth Research Laboratory funded by medatixx
GmbH & Co. KG.
Web Migration - A Survey Considering the SME Perspective
261

REFERENCES
Aversano, L., Canfora, G., Cimitile, A., and De Lucia, A.
(2001). Migrating legacy systems to the Web: an ex-
perience report. In Proc. of CSMR’01, pages 148–157.
IEEE.
Bernhart, M., Mauczka, A., Fiedler, M., Strobl, S., and
Grechenig, T. (2012). Incremental reengineering and
migration of a 40 year old airport operations system.
In Proc. of ICSM’12, pages 503–510. IEEE.
Bodhuin, T., Guardabascio, E., and Tortorella, M. (2002).
Migrating COBOL systems to the Web by using the
MVC design pattern. In Proc. of WCRE’02, pages
329–338. IEEE.
Cai, Z., Zhao, L., Wang, X., Yang, X., Qin, J., and Yin,
K. (2015). A Pattern-Based Code Transformation Ap-
proach for Cloud Application Migration. In Proc. of
CLOUD’15, pages 33–40. IEEE.
Canfora, G., Fasolino, A. R., Frattolillo, G., and Tramon-
tana, P. (2008). A wrapping approach for migrating
legacy system interactive functionalities to Service Ori-
ented Architectures. Journal of Systems and Software,
81(4):463–480.
European Commission (2003). Commission Recommen-
dation of 6 May 2003 concerning the definition of
micro, small and medium-sized enterprises (notified
under document number C(2003) 1422) (OJ L 124,
20.5.2003, pp. 3641).
Gipp, T. and Winter, A. (2007). Applying the ReMiP to Web
Site Migration. In Proc. of WSE’07, pages 9–13. IEEE.
Huang, D., Yi, L., Song, F., Yang, D., and Zhang, H. (2014).
A secure cost-effective migration of enterprise applica-
tions to the cloud. International Journal of Communi-
cation Systems, 27(12):3996–4013.
Jamshidi, P., Ahmad, A., and Pahl, C. (2013). Cloud Migra-
tion Research: A Systematic Review. IEEE Transac-
tions on Cloud Computing, 1(2):142–157.
Karampaglis, Z., Mentis, A., Rafailidis, F., Tsolakidis,
P., and Ampatzoglou, A. (2014). Secure Migration
of Legacy Applications to the Web, pages 229–243.
Springer.
Khadka, R., Reijnders, G., Saeidi, A., Jansen, S., and Hage,
J. (2011). A method engineering based legacy to SOA
migration method. In Proc. of ICSM’11, pages 163–
172. IEEE.
Khadka, R., Saeidi, A., Idu, A., Hage, J., and Jansen, S.
(2013). Legacy to SOA Evolution: A Systematic Lit-
erature Review. In Migrating Legacy Applications:
Challenges in Service Oriented Architecture and Cloud
Computing Environments, chapter 3, pages 40–71. IGI
Global.
Kienle, H. M. and Distante, D. (2014). Evolution of Web
Systems. In Evolving Software Systems, chapter 7,
pages 201–228. Springer, 1 edition.
Lewis, G., Morris, E., and Smith, D. (2006). Analyzing the
reuse potential of migrating legacy components to a
service-oriented architecture. In Proc. of CSMR’06,
pages 9 pp.–23. IEEE.
Lucia, A., Francese, R., Scanniello, G., Tortora, G., and Vi-
tiello, N. (2006). A Strategy and an Eclipse Based
Environment for the Migration of Legacy Systems
to Multi-tier Web-based Architectures. In Proc. of
ICSM’06, pages 438–447. IEEE.
Lucia, A. D., Francese, R., Scanniello, G., and Tortora, G.
(2008). Developing legacy system migration methods
and tools for technology transfer. Software: Practice
and Experience, 38(13):1333–1364.
Marchetto, A. and Ricca, F. (2008). Transforming a Java
application in an equivalent Web-services based appli-
cation: Toward a tool supported stepwise approach. In
Proc. of WSE’08, pages 27–36. IEEE.
Mohagheghi, P. and Sæther, T. (2011). Software Engineer-
ing Challenges for Migration to the Service Cloud
Paradigm: Ongoing Work in the REMICS Project. In
Proc. of SERVICES’11, pages 507–514. IEEE.
Nussbaumer, N. and Liu, X. (2013). Cloud Migration for
SMEs in a Service Oriented Approach. In Proc. of
COMPSAC’13, pages 457–462. IEEE.
Petersen, K., Vakkalanka, S., and Kuzniarz, L. (2015).
Guidelines for conducting systematic mapping studies
in software engineering: An update. Information and
Software Technology, 64:1–18.
Razavian, M. and Lago, P. (2010). A Frame of Reference for
SOA Migration. In Towards a Service-Based Internet,
volume LNCS 6481, pages 150–162. Springer.
Rose, J., Jones, M., and Furneaux, B. (2016). An integrated
model of innovation drivers for smaller software firms.
Information & Management, 53(3):307–323.
Sosa, E., Clemente, P. J., Conejero, J. M., and Rodriguez-
Echeverria, R. (2013). A model-driven process to mod-
ernize legacy web applications based on service ori-
ented architectures. In Proc. of WSE’13, pages 61–70.
IEEE.
Spanos, D.-E., Stavrou, P., and Mitrou, N. (2012). Bringing
relational databases into the Semantic Web: A survey.
Semantic Web, 3(2):169–209.
Strauch, S., Andrikopoulos, V., Bachmann, T., Karastoy-
anova, D., Passow, S., and Vukojevic-Haupt, K. (2013).
Decision Support for the Migration of the Application
Database Layer to the Cloud. In Proc. of CLOUDCOM-
ASIA’13, pages 639–646. IEEE.
Wohlin, C. (2014). Guidelines for snowballing in systematic
literature studies and a replication in software engineer-
ing. In Proc. of EASE’14, pages 1–10. ACM.
Yew Wong, K. and Aspinwall, E. (2004). Characterizing
knowledge management in the small business environ-
ment. Journal of Knowledge Management, 8(3):44–61.
Zhang, X., Shen, B., Tang, X., and Chen, W. (2010). From
isolated tenancy hosted application to multi-tenancy:
Toward a systematic migration method for web appli-
cation. In Proc. of ICSESS’10, pages 209–212. IEEE.
ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering
262