Intelligent Elearning Environments

searchcoilΛογισμικό & κατασκευή λογ/κού

15 Αυγ 2012 (πριν από 5 χρόνια και 3 μήνες)

226 εμφανίσεις

Proposal part B, page [x] of [
y]

Confidential document

Only for use by project partners


CampusSource Concept Proposal

for an application within the framework of the 2
nd

call

Working Title:

COCOON


Problem description


In view of the increasingly shorter development times for software i
nnovations, the
European software industry faces the challenge of structuring software production
more efficiently by using external inputs. One approach in this direction is offered by
open source software products (OSS). While the continuing increase in
the
significance of OSS has led to a situation where a large number of software
components are made available within the scope of open source licence conditions,
the use of open source software components in the industry is still at a relatively low
level.
1

The reasons for this low diffusion of OSS lie in the technological, economical
and legal aspects of the supply and demand sides of the OSS market.


These aspects are explained below.




When OSS products are created and used,
legal aspects
must be consider
ed in
regard to the intangible property rights of software. In particular the copyright
and patent laws that are applied on the producer side as an essential instrument
for controlling OSS diffusion are significant. On the demander side the legal
propertie
s determine the adoption of OSS within the framework of software
production processes. Legal adoption barriers can thus exist due to the lack of
legal transparency and security of OSS products.




The reuse of software regularly follows economic goals affect
ing influencing
variables such as costs, quality and the time when the primary product should
be launched on to the market. Consequently, economic barriers for using the
products arise from the specific
economic risks

that have to be considered
within the
scope of decision
-
making in regard to the use of OSS components
as secondary products. These risks should be made accessible for an economic
evaluation in the form of
transaction costs
. Transaction costs for OSS are
incurred for example when searching for
suitable components (search costs),
adapting components (adaptation costs) and coordinating software production
processes (coordination costs). Major economic adoption barriers for OSS are
triggered by the lack of models and techniques to provide a structu
red



1

Cf. Friedewald, M., Blind, K., Edler, J.: Die Innovationstätigkeit der deutschen Softwareindustrie,
(Innovation in the German Software Industry) Page 158 ff., in: Wirtschaftsinformatik, Volume 44
(2002), Edition 2, Page 151
-
161.

Proposal part B, page [x] of [
y]

evaluation of transaction costs and hence a basis for taking the decision to use
OSS products.




A basic prerequisite for the productive use of OSS is the availability of
suitable software
-
technological methods and tools to break down the
technological
barriers
and enable an integration of OSS components in
commercial software
production contexts. This includes concepts for long
-
term quality assurance and coordination of OSS components, but also
organisational and technological platforms that provide in
tegrated support for
the entire software production process. These must also provide organisational
integration in a distributed, collaborative development process which is typical
for the development and maintenance of open source software products.


Whil
e SourceForge.net is an OSS repository for the development of exchange
processes between the supply and demand sides of OSS, this platform only provides
support for source text
-
centred activities for the implementation and maintenance of
OSS projects. Ther
e is no support for analysis and design activities that would provide
a significant contribution to the structured, functional description of OSS. Focussing
on the downstream phases of the development process leads to a situation where the
informational ba
ses for the functional
-
conceptual system description are missing that
would ensure a methodically supported reuse of OSS and which could consequently
have a positive influence on the sustainable diffusion of open source software


especially in the plann
ing and design of software projects. In addition, it can be
determined that there are no mechanisms available that could break down the legal
and economic barriers and offer integrated management of all the sub
-
aspects of OSS.
The proprietarisation of the
SourceForge.net platform would also lead to process
losses that would restrict the free exchange of OSS products.

Research areas

In view of the adoption barriers that have been identified, we are faced with the task
of developing integrated methods, conce
pts and tools that could handle all the
diffusion aspects of OSS and its interdependencies and thus make a contribution
towards the sustainable dissemination of OSS components within the framework of
commercial and non
-
commercial software production contex
ts. A conceptual basis for
structuring this object area and deriving research areas is provided by the individual
activities of the software development process (cf.
Fig.
1
).

Proposal part B, page [x] of [
y]

Software Development Activities
Specific research areas
Analysis
Design
Implemen-
tation
Integration,
maintenance
Planning
1. Ontologies and domain engineering for OSS
Overlapping research areas
5. Organisational aspects of OSS: Portal management, flow and structural organisation:
distributed users, communication, collaboration and community building
6. Intelligent technologies
7. Economic valuation of OSS and business models for OSS
8. Legal aspects of OSS: Copyright and patent law regulation areas, licence models
2. Functional components and specifications,
distributed reference modelling
3. Quality assurance
4. Distributed implementation, refactorisation,
test, optimisation

Fig.
1
: Research areas

The content of these research areas will be described in the following.

Activity
-
specific research areas


1. Ontologies and domain engineering for OSS


To make well
-
founded decisions regarding the use of OSS for software developm
ent
it is necessary to have an information basis that can provide statements about the
technical suitability of available OSS components. This information basis is required
within the framework of a reuse
-
oriented development approach, especially in the
pl
anning and analysis phases. Methods and tools for developing knowledge domains
and ontologies can be used to exploit knowledge about existing OSS products.
2

At
present there are no standardised ontologies to describe existing OSS components.


Relevant res
earch topics of this area are, e.g.:




Methods and tools to generate and match metadata, knowledge domains and
ontologies




2

Cf. the Ontolo
gy Tool Project by Stanford University: Protege in http://protege.stanford.edu [2003
-
09
-
10].

Proposal part B, page [x] of [
y]



Component representations



Semantic Web and Content Management



Intelligent agents to automate and decentralise knowledge exploitation fr
om
component representations



Classification systems


2. Design and architecture


Within the scope of analysis and design, functional
-
conceptual and technical aspects
in regard to the integration of OSS in existing or evolving software products play a
deci
sive role. This includes the task of providing a description and analysis of
components for defined (functional) domains. In addition, it must be ascertained what
influence the integration of OSS has on existing software architectures and in which
way thes
e effects can be handled.


Relevant research topics in this area are e.g.:




Tools and methods for functional
-
conceptual and technical OSS component
modelling (basic components, functional components)



Support for the integration of components in architectu
res



Coordination and control of distributed modelling processes



Reference modelling for OSS components



Functional
-
conceptual modelling for defined domains (e.g. eLearning)



Prototype implementation of concepts within the scope of development
platforms (e.g
. Eclipse, NetBeans)


3. Quality assurance


The existing source code is an important property of OSS components. Thus, it is not
just possible to reuse these components in the form of a blackbox, but rather also as a
whitebox. Although empirical research w
ork has been carried out on the subject of
OSS source code,
3

to date these have only delivered partial findings for source code
analysis and have not provided any useable concepts or tools for the quality assurance
of OSS components.


Relevant research to
pics in this area are e.g.:




Quality assurance through source code analysis and languages (e.g. XML)



Prototype implementation of tools for quality assurance within the scope of
development platforms (e.g. Eclipse, NetBeans)



Derivation of models and metadat
a from OSS programs (Reverse
Engineering)


4. Distributed implementation, factorisation, test and optimisation





3

A study relevant to OSS source code analysis is "Orbiten Free Software Survey". Cf.
http://orbiten.org/ofss/01.html [2003
-
09
-
10].

Proposal part B, page [x] of [
y]

The institutional and spatial distribution of OSS development processes places
considerable design demands on the implementation and maintenance

of software
products.


Relevant research topics in this area are e.g.:




Tools for the coordination and control of distributed implementation activities



Support services for development platforms (e.g. Eclipse, NetBeans)



Tools for quality assurance in the

integration of OSS components



Methods for distributed configuration management/change management of
OSS components


Overlapping research areas


5. Organisational aspects of OSS


In regard to all software development activities, aspects of structural and p
rocedural
organisation that characterise the replacement relationships between producers and
users of OSS must be considered. Organisational instances are, for example,
mediators
, which provide specific services to support the OSS production process. All
t
he processes from OSS production to OSS use are included within the framework of
a procedural
-
organisational consideration


the interaction behaviour between the
actors (producers, mediators, users) and the software development models must also
be conside
red. In this way it is possible to create transparency in regard to process
structures which can be used as a basis for an analysis of weak points. Tools that can
evaluate the performance of the processes are required to control this process
(performance t
ransparency).


Relevant research topics in this area are e.g.:




OSS and software development models



Reference models to describe and analyse organisational instances (e.g.
mediators) and processes in OSS development



Business models for mediators


6. Intell
igent technologies


The research field of intelligent technologies covers concepts and methods aimed at
intelligent support for the software production process.


Relevant research topics in this area are e.g.:




Behaviour
-
oriented approaches and distribut
ed adaptive systems



Analysis procedures (statistical techniques, data mining)



Prototype implementation of intelligent technologies within the scope of
development platforms (e.g. Eclipse, NetBeans)


8. Economic evaluation of OSS and business models for OSS

Proposal part B, page [x] of [
y]


Relevant research topics in this area are e.g.:




Influencing factors and models for OSS diffusion and/or adoption



Transaction costs and OSS



Support for decision
-
making processes to evaluate economic properties of
OSS



Controlling of OSS projects (e.g. fea
sibility study, TCO)



Risk management for OSS projects



Business model for OSS components



Benchmarking of OSS projects


9. Legal aspects of OSS


Relevant research topics in this area are e.g.:




Modelling and implementation of legal aspects of OSS in the deve
lopment
process



Legally compliant structure of development platforms (e.g. Eclipse, Netbeans,
SourceForge)



Legal management for OSS projects



Support for decision
-
making processes to evaluate legal properties of OSS
components



"Licence paths" of OSS (e.g.
roll
-
out of existing software products as OSS)


Implementation


The bases for the design of platforms required for the problem
-
oriented support of the
open source software engineering process will be created within the framework of the
above
-
mentioned res
earch areas. For example, the results of the research can be
implemented in the form of prototypes based on available systems. For instance, the
Eclipse and NetBeans platforms are development platforms distributed under IBM’s
Common Public License (CPL) an
d Sun Microsystems’ Sun Public License (SPL).
Integration can take place, for example via plug
-
ins (cf.
Fig. 2
).


Proposal part B, page [x] of [
y]

J2ME
J2SE
GUI
Servlet / JSP
EJB
JDBC
J2EE
analysis
develop
design
test
deploy
maintain
Plug-In
manage
Eclipse
bzw.
NetBeans
Plug-In
Plug-In
Plug-In

Fig. 2: Degree of coverage of the development process by the
Eclipse
and

NetBeans
pl
atforms