Using a Domain-specific Modeling Language for Analyzing Harmonizing
and Interfering Public and Private Sector Goals
A Scenario in the Context of Open Data for Weather Forecasting
Sietse Overbeek
1
and Marijn Janssen
2
1
Institute for Computer Science and Business Information Systems, University of Duisburg-Essen,
Reckhammerweg 2, D-45141 Essen, Germany
2
Faculty of Technology, Policy and Management, Delft University of Technology,
Jaffalaan 5, 2628 BX Delft, The Netherlands
Keywords:
DSML, e-Government, GoalML, Goal Modeling, Open Data.
Abstract:
The opening of data by public organizations can result in innovations and new business models in the private
sector. Yet, the public and private sectors may have different and sometimes interfering objectives. In this
paper, we analyze the goals of an open data business model for weather forecasting using the multi-perspective
goal modelling language GoalML. The public and private sectors partly share similar goals, but creating public
value was found to be interfering (to some extent) with the private sector objective of making profit. One of
the values of GoalML is that it clearly shows harmonizing and interfering goals. The interfering goals are one
of the explanations for a slow adoption of open data. Mechanisms need to be developed to deal with them.
1 INTRODUCTION
In the past ten years, the opening of public sector
data, or open data’ for short, has gained increas-
ing attention. Open data can be defined as: “non-
privacy-restricted and non-confidential data which
is produced with public money and is made avail-
able without any restrictions on its usage or distribu-
tion” (Janssen et al., 2012, p. 258). Data excluded
from this definition concern private, confidential, and
classified data. Open data can be provided by both
public or private organizations, as such, in contrast to
the previous definition it is not necessary to be col-
lected or produced with public money (O’Riain et al.,
2012). There are at least four motivating statements
to make use of open data. First, it provides greater
return on public investments. Second, policy-makers
are provided with data needed to address complex
problems (Arzberger et al., 2004). Third, it is pos-
sible to tap into the intelligence of the crowd by en-
abling citizens to participate in analyzing large quan-
tities of data sets (Surowiecki, 2004). Fourth, orga-
nizations can improve their accountability and trans-
parency (Janssen et al., 2012; Zhang et al., 2005). The
growth of open data does not merely come with bene-
fits, as it is also known that organizations have to deal
with adoption barriers in order to make data openly
available and to let it be used successfully (Janssen
et al., 2012). Governments release their data to prop-
agate their public values, that form the basis of the
democratic system (Moore, 1995). There seems to
be no common definition of public values (Cordella
and Bonina, 2012). However, they cannot be merely
embraced by special interest groups, as public values
should be part of society as a whole (Jørgensen and
Bozeman, 2007). In other words: “the public sector
is there for everybody, it is not the extended arm of
a particular class or group” (Jørgensen and Bozeman,
2007, p. 361). The propagation of public values by
public organizations can be directly associated with
the achievement of goals that these organizationsneed
to fulfill in order to pursue public values.
One of the problems concerns the situation where
goals of private organizations that make use of pub-
lic sector data might differ in part from goals that
public organizations have in relation to that identi-
cal public sector dataset. More precisely, one of the
complexities in open data is the involvement of or-
ganizations having different goals than the organi-
zation that is the provider of the data. In this pa-
per, both public and private sector goals related to
the opening and usage of data are analyzed to un-
derstand to what extent they are similar, harmoniz-
ing, different, or even interfering with each other. It
531
Overbeek S. and Janssen M..
Using a Domain-specific Modeling Language for Analyzing Harmonizing and Interfering Public and Private Sector Goals - A Scenario in the Context of
Open Data for Weather Forecasting.
DOI: 10.5220/0005237505310538
In Proceedings of the 3rd International Conference on Model-Driven Engineering and Software Development (MODELSWARD-2015), pages 531-538
ISBN: 978-989-758-083-3
Copyright
c
2015 SCITEPRESS (Science and Technology Publications, Lda.)
can be seen that whereas the public sector is focused
on creating public values, the private sector is profit-
oriented. They will be less concerned with, for ex-
ample, ensuring social security or increasing citizen
empowerment, although they need to comply with
governmental regulations. The goal analysis will be
conducted in the context of a ‘Weather Radar’ sce-
nario, which is inspired by a weather forecasting ser-
vice operated by a Dutch private organization called
‘Buienradar’ (www.buienradar.nl), which makes use
of open data to provide real-time weather information
for their clients. Competitors of ‘Buienradar’ include
Druppel (www.druppel.nu), Shower Alarm (‘Buien-
alarm’ in Dutch, see: www.buienalarm.nl), and Drash
(dra.sh). The weather data is collected by a semi-
public meteorological organization funded by pub-
lic money. In case of Buienradar, data is collected
from the Royal Netherlands Meteorological Institute,
of which the abbreviation of the Dutch translation
is ‘KNMI’ (www.knmi.nl). These (semi-)public and
private organizationshave goals related to the opening
and use of open data and, therefore, provide a proper
context on which our scenario is inspired.
The multi-perspective Goal Modelling Language
(GoalML) (Overbeek et al., 2015) is used to design
goal models for the Weather Radar case from the
point of view of the involved organizations. It is
shown that these goal models enable to perform dif-
ferent kinds of analyses, which includes but is not
limited to: determination of how goals are ordered
in a goal hierarchy, which goals are similar and har-
monizing, or which goals are different and even in-
terfering. There are three main reasons why GoalML
has been selected for the design of the goal models.
First, GoalML models are an integral part of enter-
prise models, which provide relevant contexts, such
as: Descriptions of resources, business process mod-
els or models of the IT infrastructure (Overbeek et al.,
2015; Frank, 2014). Second, while it is possible to
model goals with a general purpose modelling lan-
guage (GPML) like the Unified Modelling Language
(UML) or the Entity-Relationship Modelling (ERM)
language, the GoalML is actually a domain-specific
modelling language (DSML). This is for three rea-
sons: Using a GPML requires a modeler to recon-
struct relevant concepts such as various kinds of goals
from scratch, which compromises modelling produc-
tivity. Furthermore, a DSML includes specific con-
straints that prevent modelers to a certain degree from
creating erroneous models. Finally, a DSML enables
the use of a specific visual notation or, in other words,
a concrete syntax, which fosters comprehensibility.
The structure of this paper is as follows. Section 2
describes the background of the scenario that is used
as a basis for the creation of goal models related to
the opening and using of weather data. The goal mod-
els are presented in section 3 and, subsequently, dis-
cussed in section 4. The paper ends with conclusions
and future research in section 5.
2 A SCENARIO FOR OPENING
AND USING WEATHER DATA
The motivation to stimulate organizations in opening
their data is embodied in the European Union (EU)
Public Sector Information (PSI) directive, which was
released in 2003 (EU, 2003). This directive is based
on two ideas. First, public sector data should be
made available for third parties at low prices and un-
restrictive conditions, and, second, this would ensure
a ‘level playing field’ among organizations, which
means that equal opportunities are provided for or-
ganizations. One of the objectives of the publica-
tion of open data is to facilitate the innovative use
of these data by companies (Dawes, 2010; O’Riain
et al., 2012; Neuroni et al., 2013). The European
Commissioner believes that open data boosts the Eu-
ropean economy by e40 billion per year (EC, 2010).
These prospects of the publication of open data lead-
ing to possible usage by third parties directly relate to
the development that organizations increasingly use
social media to facilitate interactions between them-
selves and their clients (Chun et al., 2012). Social
media are considered to be “a group of Internet-based
applications [...] that allow [for] the creation and ex-
change of user-generated content” (Kaplan and Haen-
lein, 2010, p. 61). The Weather Radar weather fore-
casting service uses open data collected with public
money and after enriching the data it is provided to
the users. They employ two channels to interact with
their users, which includes a Web site and an applica-
tion, or app for short, which can be downloaded and
installed on mobile devices.
The combination of open data and social media
has led to the introduction of so-called infomediary
business models, which “can be initiated by [...] pub-
lic or private [organizations] and are aimed at sup-
porting the coordination between open data providers
and users” (Janssen and Zuiderwijk, 2014, p. 2). A
business model in general contains the rationale and
the elements required to accomplish certain organiza-
tional objectives (Keen and Qureshi, 2006). The rev-
enue models of the Web site and the app are slightly
different, as the Web site primarily depends on adver-
tisements, whereas the app provides advertisements
and options to buy additional content within the app
itself. For example, this includes the options to buy 3
MODELSWARD2015-3rdInternationalConferenceonModel-DrivenEngineeringandSoftwareDevelopment
532
Data provider
Meteorological Institute
Users
(citizens, businesses, public
organizations, researchers, …)
Weather Radar
(website, app)
Figure 1: Open data network for the Weather Radar scenario.
hours or 24 hours rain forecasts, information about
thunder, hail and sun power. The app is based on
a so-called ‘single-purpose app’ infomediary busi-
ness model (Janssen and Zuiderwijk, 2014). “Single-
purpose apps provide real-time services such as infor-
mation about weather, quality of restrooms, vehicles,
houses, and pollution. These apps often provide a sin-
gle function, based on one type of provided open data.
The app processes the data and presents it visually
for the ease of the users” (Janssen and Zuiderwijk,
2014, p. 11). The ‘open data network’ for this sce-
nario is shown in figure 1. The data is primarily based
on information collected by a meteorological insti-
tute, which is the semi-public open data provider. The
private organization offering the weather forecasting
service extrapolates the data to make predictions and
visualizes the data on a geographical map. For users,
this is complicated to realize by themselves and as
such there are hardly any users who use this raw open
data correctly (Janssen and Zuiderwijk, 2014). The
line at the top of figure 1 indicates that it is, however,
possible to use raw open data directly without relying
on an infomediary as a liaison party.
3 GOALS FOR OPENING AND
USING DATA IN THE
SCENARIO
Figure 2 presents the goal model of a meteorological
institute being the semi-public sector data provider.
This goal model includes the goals that this institute
wants to achieve by opening up their weather data to
be used by third party organizations such as infome-
diary organizations providing a weather forecasting
service. Figure 3 shows the goal model of an infome-
diary that uses open weather data to deliver a weather
forecasting service to possible users. The goal models
have been designed by using GoalML. The diagrams
include both so-called engagement goals and sym-
bolic goals (Overbeek et al., 2015). An engagement
goal is a goal of which the desired result is quantifi-
able, for example, the goal ‘increase citizen services’
shown in figure 2 is an engagement goal, as it is quan-
tifiable whether the number of new services have in-
deed increased. Motivations and the performance of
employees responsible for goal achievement can also
be made more explicit by means of such engagement
goals. An engagement goal is visualized as a target.
A symbolic goal is a goal of which the desired result
is not directly quantifiable and includes a qualitative
aspect and is visualized as a lighthouse. An example
of a symbolic goal found in figure 2 is increase citi-
zen satisfaction’, as the increase of citizen satisfaction
is not directly quantifiable. A symbolic goal ‘increase
trust’ is shown on top of figure 2. The star symbol
with the number one inside the star shows that this
specific goal has the highest priority.
The circles shown on the top right of each of the
goals depict specific goal matter, further specifying
the goal content. A yellow hexagon with a plus sym-
bol in it is part of the goal matter of the mentioned
symbolic goal. This shows that something needs to
increase upon achievement of the goal, in this case
the trust of the semi-public agency. In contrast to the
hexagon with a plus symbol, a hexagon with a minus
symbol indicates that something needs to decrease.
The symbol of an eye looking at a diamond as part
of the goal matter of the symbolic goal shows that
the goal content is qualitative in nature. When fur-
ther interpreting the diagrams shown in figures 2 and
3, it can be determined that two other symbols can be
part of the goal matter, which are the indicator sym-
bol and the ‘object’ symbol. For example, the goal
matter of the ‘increase added value’ goal shown in
figure 3 contains an indicator symbol, expressing that
the goal content is quantitative in nature. The ‘ob-
ject’ symbol is illustrated by means of a combination
of a circle, a triangle, and a rectangle. This symbol
is used to indicate that an explicit ‘object’ is part of
the goal content. For example, the ‘object’ symbol
in the goal matter of the ‘keep existing users’ goal as
part of figure 3 indicates that a ‘user’ is a specific ob-
ject to take into account as part of the goal content.
Next to the two different kinds of goals and the goal
content, there are three kinds of relationships that can
be found in both diagrams: A causal relationship, a
meansend relationship, and a mathematical relation-
ship. The causal relationships are indicated by means
of domino pieces, together with an arrow that points
in the upward direction indicating a positive causal
relationship. For example, the goal ‘increase citizen
satisfaction’ found in figure 2 has a positive causal
UsingaDomain-specificModelingLanguageforAnalyzingHarmonizingandInterferingPublicandPrivateSectorGoals-
AScenariointheContextofOpenDataforWeatherForecasting
533
Increase
transparency
1
Transparency
Increase
accountability
1
Accountability
Increase trust
1
Trust
Increase
citizen
satisfaction
1
Citizen
satisfaction
Increase
citizen
empowerment
3
Citizen
empowerment
Increase data
provider
visibility
4
Data provider
visibility
Increase data
scrutinization
6
Data
scrutinization
Increase
citizen
services
Citizen services
Increase
quality of
citizen
services
2
Quality of citizen
services
Keep data
access equal
5
Data access
Increase
quality of
policy-making
processes
5
Quality of policy-
making
processes
Increase new
insights in
public sector
7
New insights in
public sector
Increase
stimulation of
knowledge
development
8
Stimulation of
knowledge
development
3
Figure 2: Goal model of a semi-public meteorological institute for opening weather data.
MODELSWARD2015-3rdInternationalConferenceonModel-DrivenEngineeringandSoftwareDevelopment
534
Increase
continuity
1
Continuity
Increase profit
2
Profit
Increase
added value
4
Added value
Decrease
data access
costs
6
Data access
costs
Increase
innovation
4
Innovation
Increase
efficiency
5
Efficiency
Increase
service quality
4
Service quality
Increase
differentiation
5
Differentiation
Increase
users
Users
3
Increase data
quality
6
Data quality
Keep existing
users
3
Users
Figure 3: Goal model of an infomediary organization that uses open weather data.
effect on the increase trust goal’. The meansend re-
lationships are combined with the causal (and other)
relationships shown in the diagrams. For example, the
goals ‘keep existing users’, ‘increase added value’,
and ‘increase service quality’ shown in figure 3 are a
means to support in reaching another final goal, which
is the ‘increase users’ goal in this case. Finally, the
positive mathematical relationship between, for ex-
ample, the goals ‘increase efficiency’ and ‘increase
profit’ indicates that increasing the efficiency has a
positive mathematical effect on the actual profit that
is made by the infomediary.
4 DISCUSSION
When comparing both diagrams, there are differences
and interferences that can be identified. Fortunately,
there are also similarities and harmonious circum-
stances that can be identified. First, identified dif-
ferences and interferences will be described. Sec-
ond, the similarities and harmonious circumstances
are discussed. Obviously, it can be seen that the goal
model of the semi-public agency for opening weather
data contains more goals and the goal hierarchy itself
is deeper. Consequently, for a semi-public agency like
a meteorological institute the structure of lower-level
goals that need to be achieved as a result of opening
up weather data is more complex and it seems more
demanding to achieve the topmost goal in the hierar-
chy. This has to do with the fragmented government
structure in which different organizations have differ-
ent priorities (Kraaijenbrink, 2002). For example, the
Dutch Ministry of Economics focuses on value cre-
ation and innovation by businesses. The Dutch Min-
istry of Interior has prioritized goals related to trans-
parency, reputation, and improved democracy (Plas-
terk, 2014). For the meteorological institute as part of
the scenario as discussed in section 2, trust, account-
ability, and citizen satisfaction are high on the agenda
as well. This means that the opening of data requires
prioritizing goals and making trade-offs. When in-
terpreting the goals that have the highest priorities,
there seem to be big differences between the open
data provider and the infomediary. The top priorities
for the data provider have to do with increasing trust,
transparency, accountability, and citizen satisfaction
by opening data, while the top priority for the info-
mediary is to increase its continuity by using open
data. For the meteorological institute being the data
provider, an increase in trust can be achieved after
achieving an increase in transparency, accountability,
UsingaDomain-specificModelingLanguageforAnalyzingHarmonizingandInterferingPublicandPrivateSectorGoals-
AScenariointheContextofOpenDataforWeatherForecasting
535
Increase
innovation
4
Increase
efficiency
5
Increase
users
3
Increase data
quality
6
Keep existing
users
3
Increase
citizen
satisfaction
1
Increase
citizen
empowerment
3
Increase data
scrutinization
6
Increase
citizen
services
Keep data
access equal
5
Increase new
insights in
public sector
7
Increase
stimulation of
knowledge
development
8
3
Increase profit
2
Figure 4: Harmonizing and interfering goals related to open weather data.
and citizen satisfaction. When interpreting this goal,
however, it is assumed that an ‘increase in trust’ can
be interpreted differently by those who are responsi-
ble to achieve this goal. People can come up with
different interpretations in case they want to deter-
mine when a governmentagency can be ‘trusted’. The
term trust is somehow related to terms like hope, con-
fidence, belief, and commitment and deals with ac-
tively anticipating and facing an unknown future (Sz-
tompka, 1999, p. 25). When acting in uncertain
and uncontrollable conditions, people take risks and
make bets about the future. As such, this key goal of
the open data provider might be understood in many
different ways, which makes the achievement of this
goal seemingly more complex than achieving the key
goal(s) of the infomediary.
For the infomediary, the achievement of an in-
crease in continuity directly relates to the achieve-
ment of an increase in profit. The goal of increas-
ing their profit is reflected in their business model,
as advertisements and options to buy content within
the mobile app are provided. The goal to increase
profit that is related to the private value to increase
profit might interfere with the public value to in-
crease citizen satisfaction. Analogous to how soft-
ware is developed under the GNU General Public Li-
cense (see: www.gnu.org/licenses/gpl-3.0.en.html), a
government agency that offers open data wants end
users to use, share, and extrapolate this data for free.
Users can get bothered by advertisements while us-
ing the weather forecasting service and they might
not want to be bothered with additional content that
must be bought to get access to more advanced fea-
tures or additional data. Especially when the mobile
app allows users to pay for features that allow ac-
cess to additional data, there is interference between
the ‘increase profit goal’ of the infomediary and the
goal ‘keep data access equal’ of the semi-public data
provider. As can be seen in figure 2, when this goal
cannot be achieved, the possibility to achieve the re-
lated goals that are higher in the hierarchy, such as
‘increase citizen empowerment’ and ‘increase citizen
satisfaction’ is threatened as well. Another side-effect
of providing the users the possibility to pay additional
content within the app is that users might become dis-
appointed after realizing they have to pay for addi-
tional features or content after having downloaded an
app that was downloadable at zero costs. This also
threatens the goal of the semi-public data provider
to increase citizen satisfaction. Although there are
differences and interferences to be found when com-
paring the key goals of both models, an increase in
the trust level of the data provider presumably con-
tributes to achievement of the ‘increase continuity’
goal of the infomediary. It is assumed that the con-
tinuity of an infomediary is easier to protect in case
MODELSWARD2015-3rdInternationalConferenceonModel-DrivenEngineeringandSoftwareDevelopment
536
the used open data originates from a trustworthy data
provider. However, differentiating from other com-
petitors in the market and asking citizens to pay com-
mercial rates in order to reach some higher-order goal
will never fit with public sector goals, as these kind
of goals are not concerned with commercial interests.
Another harmonizing aspect found in the two goal
models is related to the users of the weather forecast-
ing service. Achieving the goal ‘increase citizen sat-
isfaction’ found in the goal model of the meteorolog-
ical institute will probably have a positive effect on
the goal ‘keep existing users’. The assumption un-
derlying this statement is that once the satisfaction
of citizens that are also existing users of the weather
forecasting service increases, the probability they will
keep using this service increases as well.
The goals ‘increase data scrutinization’ as part of
the goal model of the meteorological institute and ‘in-
crease data quality’ as part of the infomediary goal
model are also considered to be in harmony with each
other. If the meteorological institute increases the
possibilities to scrutinize its open data, this might lead
to a further improvement of the data quality. The goal
‘decrease data costs’ is considered to be in harmony
with the goal ‘keep data access equal’. The goals
‘increase stimulation of knowledge development’ and
‘increase new insights in public sector’ are in har-
mony with the goal ‘increase innovation’. The in-
fomediary might benefit from an increase in new in-
sights in the public sector which can lead to an in-
crease in innovation for the infomediary and, vice
versa, the open data provider might benefit from in-
novation in the private sector to gain new insights
that are relevant for them. Figure 4 shows the result-
ing goal model in which it has been visualized which
goals are in harmony with each other and which goals
are interfering with each other. Goals that have rela-
tionships with a symbol of a sun attached to it show
that these goals are in harmony with each other. The
symbol with a dark cloud and a lightning flash indi-
cates that goals are interfering. The priority symbols
of those goals that belong to the open data provider
have been colored green, while the priority symbols
attached to the goals belonging to the infomediary
have a red color. Note that although there are goals
shown in figure 4 that are harmonizing or interfering,
this does not imply that this situation should be identi-
cal for other open (weather) data scenarios. Whether
goals are in harmony or interfering also depends on
which measurements are taken to achieve goals. For
example, there might be ways to increase profit with-
out interfering with the goal ‘increase citizen satis-
faction’, or, the other way around, there might also
be new insights created in the public sector that might
not boost innovation in the private sector.
5 CONCLUSIONS AND FUTURE
RESEARCH
In the past decade, the opening of public sector data
steadily but slowly grew in popularity. One of the
complexities in open data concerned the difference
in goals of those organizations involved in the open-
ing and usage of the data. In this paper, a scenario
has been presented that involved a semi-public mete-
orological institute, which released weather data for
the public. The scenario also included an infomedi-
ary, a private organization that used and extrapolated
this open weather data in order to offer an advanced
weather forecasting service to its users. Both pub-
lic and private sector goals related to the usage of
open data have been analyzed to understand to what
extent goals related to open data in the public and
private sectors are similar, harmonizing, different, or
even interfering with each other. GoalML is a DSML
that is suitable for modelling goals in the public and
private sectors and can be used to support organiza-
tions with developing, using, and maintaining goal
models. The priorities that public organizations as-
sign to their goals might be different per organization
and is dependent of their overall role within the pub-
lic administration. Already within the public admin-
istration goals are diverse and might be interfering.
Furthermore, the analysis shows clearly the conflict
of interest between the private and the public sec-
tor. In particular, it is shown that the private sector
goal to increase profit might interfere with the pub-
lic sector goal to increase citizen satisfaction and the
goal ‘keep data access equal’. The interfering goals
might be one of the reasons why the realization of
open data by public organizationscan be problematic.
Due to requirements on open data usage, private orga-
nizations might think that that sometimes interfering
goals result in choices that might not favor their inter-
est. However, it are not merely the public and private
sector goals related to open data that result in inter-
ferences, but the implementation of measurements to
achieve these goals influences relationships between
public and private sector goals as well. It might be
that mechanisms can be developed that are accept-
able for both the public and private sectors. As such,
we recommend public and private sectors to start dis-
cussing the release and use of open data and come
up with mechanisms that can satisfy the requirements
of both sectors. This recommendation is also inher-
ent to future research. GoalML offers stakeholders of
open data in the public and private sectors the pos-
UsingaDomain-specificModelingLanguageforAnalyzingHarmonizingandInterferingPublicandPrivateSectorGoals-
AScenariointheContextofOpenDataforWeatherForecasting
537
sibility to act as an instrument in further analyzing
and discussing models and scenarios to satisfy intra-
sectorial requirements. An important task to perform
in this context is to make GoalML even more suit-
able for use in different contexts such as goal analysis
of ‘open data networks’, as GoalML is rich in detail
which makes it suitable for advanced users but less
suitable for more novice users at first. Another part of
future research deals with the question how goal mod-
els can be integrated into software with the intention
to provide different forms of computer-based support
for, e.g., strategy formulation, goal achievement, and
enterprise management in general. A possible form of
computer-based support is to use the information pre-
sented in goal models for deductive purposes. For ex-
ample, information from goal models provide a foun-
dation for the generation of rules that need to be ad-
hered to when conducting tasks or processes that are
related to a goal. By adhering to these rules, it is pos-
sible to steer in a direction that would lead to goal
achievement. This part of future research also relates
to creating models at runtime and self-adaptive sys-
tems, which implies that a system adapts its structure,
functions, or processes to a (manually) modified goal
system. A self-adaptive system might also modify the
goal system itself to better cope with a changing en-
vironment.
REFERENCES
Arzberger, P., Schroeder, P., Beaulieu, A., Bowker, G.,
Casey, K., Laaksonen, L., Moorman, D., Uhlir, P.,
and Wouters, P. (2004). An international framework
to promote access to data. Science, 303(5665):1777–
1778.
Chun, S., Luna-Reyes, L., and Sandoval-Almaz´an, R.
(2012). Collaborative e-government. Transforming
Government: People, Process and Policy, 6(1):5–12.
Cordella, A. and Bonina, C. (2012). A public value per-
spective for ict enabled public sector reforms: A theo-
retical reflection. Government Information Quarterly,
29(4):512–520.
Dawes, S. (2010). Stewardship and usefulness: Policy prin-
ciples for information-based transparency. Govern-
ment Information Quarterly, 27(4):377–383.
EC (2010). Riding the wave: How Europe can gain form
the rising tide of scientific data. Final report of the
High Level Expert Group on Scientific Data, Euro-
pean Commission, Brussels, Belgium, EU.
EU (2003). Directive 2003/98/EC of 17 november 2003 on
the re-use of public sector information. Official Jour-
nal of the European Union L 345/90, The European
Parliament and the Council of the European Union,
Brussels, Belgium, EU.
Frank, U. (2014). Multi-perspective enterprise model-
ing: Foundational concepts, prospects and future re-
search challenges. Software and Systems Modeling,
13(3):941–962.
Janssen, M., Charalabidis, Y., and Zuiderwijk, A. (2012).
Benefits, adoption barriers and myths of open data and
open government. Information Systems Management,
29(4):258–268.
Janssen, M. and Zuiderwijk, A. (2014). Infomediary busi-
ness models for connecting open data providers and
users. Social Science Computer Review, 32(5):694–
711.
Jørgensen, T. and Bozeman, B. (2007). Public values: An
inventory. Administration & Society, 39(3):354–381.
Kaplan, A. and Haenlein, M. (2010). Users of the world,
unite! The challenges and opportunities of social me-
dia. Business Horizons, 53(1):59–68.
Keen, P. and Qureshi, S. (2006). Organizational transforma-
tion through business models: A framework for busi-
ness model design. In Proceedings of the 39th Hawaii
International Conference on System Sciences (HICSS
2006), pages 1–10. Computer Society Press.
Kraaijenbrink, J. (2002). Centralization revisited? Prob-
lems on implementing integrated service delivery in
the Netherlands. In Traunm¨uller, R. and Lenk, K.,
editors, Electronic Government: First International
Conference, EGOV 2002 Aix-en-Provence, France,
September 26, 2002 Proceedings, volume 2456 of
Lecture Notes in Computer Science, pages 10–17.
Springer, Berlin, Germany, EU.
Moore, M. (1995). Creating Public Value: Strategic Man-
agement in Government. Harvard University Press,
Cambridge, MA.
Neuroni, A., Riedl, R., and Brugger, J. (2013). Swiss ex-
ecutive authorities on open government data pol-
icy making beyond transparency and participation. In
Proceedings of the 46th Hawaii International Confer-
ence on System Sciences (HICSS 2013), pages 1911–
1920. Computer Society Press.
O’Riain, S., Curry, E., and Harth, A. (2012). XBRL and
open data for global financial ecosystems: A linked
data approach. International Journal of Accounting
Information Systems, 13(2):141–162.
Overbeek, S., Frank, U., and K¨ohling, C. (2015). A lan-
guage for multi-perspective goal modelling: Chal-
lenges, requirements and solutions. Computer Stan-
dards & Interfaces, 38:1–16.
Plasterk, R. (2014). Reactie op ARK Trendrapport Open
Data. Kamerbrief 2014-0000486923, Ministry of the
Interior, the Hague, the Netherlands, EU. In Dutch.
Surowiecki, J. (2004). The Wisdom of Crowds: Why the
Many are Smarter than the Few and How Collective
Wisdom Shapes Business Economies, Societies and
Nations. Doubleday, New York, NY.
Sztompka, P. (1999). Trust: A Sociological Theory. Cam-
bridge University Press, New York, NY.
Zhang, J., Dawes, S., and Sarkis, J. (2005). Exploring stake-
holders’ expectations of the benefits and barriers of e-
government knowledge sharing. Journal of Enterprise
Information Management, 18(5):548–567.
MODELSWARD2015-3rdInternationalConferenceonModel-DrivenEngineeringandSoftwareDevelopment
538