VARIANT LOGIC META-DATA MANAGEMENT FOR MODEL

DRIVEN APPLICATIONS

Allows Unlimited End User Configuration and Customisation of All Meta-data EIS

Application Features

Jon Davis and Elizabeth Chang

Curtin University of Technology, Curtin Business School, Bentley, 6102, Australia

Keywords: Meta-data, Meta-model, Variant, customisation, Logic, EIS, Lifecycle, Version control, Version

management.

Abstract: The scope for end users to influence the design and functionality of off the shelf Enterprise Information

system (EIS) applications is usually minimal, requiring pursuing expensive vendor supported

customisations. Our ongoing development of temporal meta-data EIS applications seeks to overcome these

issues, through modelling rather than coding, and with the meta-data model supporting the capability for

end users to define their own application logic meta-data, to supplement or replace the originating vendor’s

pre-defined application logic, as what we term Variant Logic. Variant Logic can be applied to any object

defined in a meta-data EIS application, and can be defined by any authorised user, without the need for

additional coding, and is available for immediate execution by the framework runtime engine. Variant Logic

is also preserved during automated meta-data application updates.

1 INTRODUCTION

Almost every application that is in practical use is

the result of hard coded program logic that has been

compiled and deployed for use as part of a

developer’s release schedule with minimal scope for

end users to influence the design and functionality of

the application without expensive customisations.

Our ongoing development of a temporal meta-

data framework for EIS applications seeks to

overcome these issues, as an example of the model

driven engineering paradigm. A meta-data EIS

application is defined and stored as a model, without

the need for additional coding, for direct execution

by an associated runtime engine.

A specific objective of the framework is to also

provide the capability for end users to define and

create their own application logic, to supplement or

replace a vendor’s pre-defined application logic, as

what we term Variant Logic, to become an alternate

variation of the application logic.

Variant Logic can be applied to any object

defined in a meta-data EIS application whether;

visual objects, logical processing objects, or as data

structures. Variant Logic can be defined by any

authorised user and can be executed by any user as

an alternative to the original application logic, and

without the need for additional coding as per the

design objectives of the meta-data EIS application.

The logic changes will also be available for

immediate execution by the framework runtime

engine, and are preserved during automated meta-

data application updates that may be provided by the

originating vendor or developer.

This paper reviews related works, examines the

application to a meta-data based application model,

and provides examples where the Variant Logic can

be used effectively in real enterprises.

2 RELATED WORKS

The following related issues have guided this

research.

2.1 Object Polymorphism

The polymorphism aspect of object oriented

programming has provided a helpful simplification

395

Davis J. and Chang E..

VARIANT LOGIC META-DATA MANAGEMENT FOR MODEL DRIVEN APPLICATIONS - Allows Unlimited End User Conﬁguration and Customisation of

All Meta-data EIS Application Features.

DOI: 10.5220/0003467303950400

In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 395-400

ISBN: 978-989-8425-56-0

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

to logic localisation and streamlined coding

practices (Dettmer, 1995).

An analogy in the meta-data EIS application are

multiple instances of the same meta-data object each

defined with different features as instances of

Variant Logic.

2.2 OMG, MDA and Reusable Objects

The aim of the Object Management Group (OMG)

with their Model Driven Architecture (MDA)

initiative is to “separate business and application

logic from underlying platform technology” (OMG,

2010).

UML is a widely adopted standard for software

development but is a semi-formal language which

lacks the precisely defined constructs to fully define

application logic (Mostafa, 2007).

Complementary strategies aim for more

plug’n’play style solutions where components are

constructed to an accessible interface standard which

allows the component objects to more readily

interact with minimal recoding (Talevski, 2003).

In (Yan, 2009) support for multiple simultaneous

versions of an application would be provided by a

hybrid interim data schema evolution that satisfies

both the current and proposed application

functionality as a stepping stone to the final version

and schema.

The temporal meta-data framework for EIS

applications allows multiple duplicate objects to

exist with separate logical definitions that can be

selected for execution.

2.3 User Configuration

While varying in complexity (Rajkovic, 2010), the

generally available configurable content for end

users tends to be limited to simplistic features, with

application customisation being required for more

advanced features. (Hagen, 1994) identified the need

for focussing on more configurable software to

benefit users and developers as a joint initiative of

software development to merge configuration and

customisation aspects.

Any feature of the meta-data EIS application can

be optionally configured by authorised users.

2.4 User Customisation

Whilst customisations may have a suitable overall

business case for an organisation to take this option,

it is often expensive and time consuming. (Dittrich,

2009).

Customisations to EIS systems require

developers fluent in the development languages and

in the detailed structure of the application logic and

structure and depending on the scale can become

significant software engineering exercises. (Hui,

2003) provides a model to optimise capturing the

requirements for customisations.

All features of the meta-data EIS application can

be defined as customised by authorised end users,

aiming at the knowledgeable business user rather

than the technical expert.

2.5 Model Driven Engineering

Alternatives to the common process of hard coded

application logic are provided by ongoing Model

Driven Engineering (MDE) which is a generic term

for software development that involves the creation

of an abstract model and how it is transformed to a

working implementation (Schmidt, 2006).

A significant proportion of the works to date

have involved modelling which contributes more

directly to streamlining code generation, processes

that are directly aimed for and dependent on highly

technical programmers such as (Ortiz, 2010) who

specifies alternate aspects and (Cicchetti, 2007) who

identify model insertion points for code insertion.

(Fabra, 2008) base their works on the UML 2

specification to seek to reduce coding and transform

models of business processes into executable forms.

(Zhu, 2009) takes a strong model generation

approach that modifies individual application

instances for each user.

In (France, 2007) they argue for future MDE

research to focus on runtime models, where these

executing models can also be used to modify the

models in a controlled manner. Such a direction

provides not only more manageable change control,

but also necessarily shifts the target of the change

agent closer to the knowledgeable business end user,

rather than relying solely on the technical

programmer, a similar goal of our temporal meta-

model framework for EIS applications (Davis,

2004).

3 VARIANT LOGIC META-DATA

MANAGEMENT

Our ongoing development of a temporal meta-data

framework for EIS applications seeks to remove the

need for hard coding by technical developers (other

than in the creation of the runtime engine and meta-

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

396

data editors), and transform the responsibility of

defining application logic to business analysts,

knowledge engineers and even business end users.

Once the meta-data EIS application logic has

been defined as the meta-data and stored as a model,

without the need for additional coding, the model is

used for direct execution by the runtime engine.

The model also supports a specific objective of

the framework which is to provide the capability for

end users to define and create their own application

logic, to supplement or replace a vendor’s pre-

defined application logic, as what we term Variant

Logic, to become a variation of the application logic

for their own specific purpose.

Variant Logic can be defined for each of the

three common layers of EIS application design:

• Database: The definitions for data storage,

management and workflow in the meta-data EIS

are defined in meta-data, hence the authorised

user is also able to define additional data entities

and attributes that can be associated with the

existing defined meta-data EIS application data.

• User Interface: The meta-data definitions for

user interfaces can also be modified and defined

with additional application features to operate

with, enhance or optionally replace existing

application functionality (Davis, 2005). At the

personal user level (single or permitted group

access) changes can be made to the application

meta-data that have not been flagged as core or

mandatory by the meta-data EIS application’s

higher level designers. Within this scope logic

definers can, within their authorisation limits;

o Remove (not delete) non-mandatory

features,

o Relocate any features between user

interface locations,

o Modify non-mandatory features,

o Define new features to support new or

existing data stores.

• Logical Processing: Data manipulation,

workflow and processing features are provided

by logical functions that are defined for

processing data, similar to many of the

functions in say Microsoft Excel. The functions

in the meta-data EIS application are used:

o As individual or compound functions,

o To provide individual processing,

o As inline or to be a user-defined

function that can be used throughout

the meta-data EIS application,

o To modify the display, value of or

storage of data,

o To perform an evaluation to be used

for logical workflow execution.

Variant Logic can be defined by any authorised

user and can be executed by any user as an

alternative to the standard application logic, subject

to the defined security authorisations. The logic

changes will also be available for immediate

execution by the framework runtime engine.

All Variant Logic is also preserved during

automated meta-data application updates that may

be provided by the originating vendor or developer.

Any logical conflicts that may arise during the

update can be precisely identified to minimise any

re-definition changes that may be required to the

user meta-data customisations to avoid wholesale re-

engineering efforts.

3.1 Authorisation Framework

In our meta-data EIS application model there are

various levels of hierarchical authorisation that can

be defined as depicted in Figure 1.

Figure 1: Class diagram of the Logic Definer Access

component of the meta-data EIS application model.

These functional authorisation processes are

governed by the following principles:

• All original meta-data is created by the

identified core logic definer and represents the

highest level of authorisation for that meta-data

S10_Logic_Definer_

Authorisation_Level

...

S11_Logic_Definer_Role

G02_Index_GUID_Reference_for

_all_Meta_Data_Objects

S02_Appli cation_Security

Rol e

...

E25_Logic_Variant

...

S12_Permitted_Variant_

Access

...

S13_Permitted_Variant

_to_Meta_Data_Aspect

...

E22_User_in_a_Role

...

E24_Timed_Access

...

VARIANT LOGIC META-DATA MANAGEMENT FOR MODEL DRIVEN APPLICATIONS - Allows Unlimited End

User Configuration and Customisation of All Meta-data EIS Application Features

397

EIS application.

• Additional logic owners can be defined as

(usually) lower level authorisations.

• Meta-data created by one logic definer cannot

be modified by a different logic definer, to

ensure application semantic integrity.

• Any logic definer can define new meta-data,

reference and invoke meta-data owned by other

logic definers (where authorised), and modify

undefined meta-data attributes of meta-data that

was created by other logic definers where this

functionality has not been restricted by the

creating logic definer.

• Meta-data defined by a higher level logic

definer always over-rides any other identical

meta-data definition created by a lower-level

logic definer – this may also occur during an

automated meta-data application update.

This dynamic and distributed editing feature of

the meta-data EIS is drastically different from the

traditional development lifecycle, further extending

the provision of genuine real-time and distributed

rapid application development capability, and

greatly reducing the incidence of locked down or

closed EIS application eco-systems that may be

restricted to vendor only modification.

The Logic Definer Access (see Figure 1) uses the

following classes to model the definition of the

change access:

• Logic Definer Authorisation Level: is a simple

hierarchy list of authorisation levels where a

higher level of authorisation always has a higher

priority and authorisation over all lower levels.

• Logic Definer Role: are the individual groups

or roles that can be assigned to designate an

identified group of functional logic definers.

• Logic Variant: is a designated identifier to

group all of the logic changes together into a

practical set as an instance of Variant Logic.

• Permitted Variant Access: identifies the meta-

data objects that the Logic Definer Role has

access to change as Variant Logic.

• Permitted Variant to Meta Data Aspect:

identifies which aspects of the meta-data for

that object can be changed as Variant Logic for

that Logic Definer Role. Meta-Data Aspects are

the internal groups of meta-data for each object

and the effect of the change ranges from minor

through to major aspects. Examples of meta-

data aspect changes are: text, colours, help,

sizes, position, alignment, access, type,

function, assignment, validation.

• Index GUID Reference for all Meta-Data

Objects: is a combined global register of the

identifier for all instances of the defined meta-

data objects (visual and non-visual).

• Application Security Role: is the list of the

available application roles.

• User in a Role: is the list where Users are

assigned to Application Security Roles and/or

Logic Definer Roles.

• Timed Access: is used to define periods of

allowed or denied access.

Some aspects of meta-data objects can be

defined to be restricted from permitting any Variant

Access, which would be applied to semantically

critical application logic components.

3.2 Defined Variant Logic

All of the meta-data that defines the application

logic is stored in the meta-data model, original and

Variant Logic.

The key aspect that differentiates the meta-data

is the identified and assigned Variant Logic for each

set of additional meta-data (see Figure 2), as the

Variant Logic identifies a specific group of meta-

data for a specific defined purpose by its Logic

Definer.

Figure 2: Class diagram of the generic assigned Logic

Variant for all objects of the meta-data EIS application

model.

3.3 Selection and Execution

As Variant Logic is defined, the new objects become

part of the overall application pool of objects and

thus are subject to the same security access

mechanisms in order to provide access to the objects

and how the runtime engine will manage the

ongoing access to the objects.

Figure 3 provides an extract of the design for the

Variant Logic Access model. It allows for assigning

access to; users, roles or on an application wide

basis.

X01_Example_General_

Meta_Data_Entity

...

E25_Logic_Variant

...

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

398

Figure 3: Class diagram of the Variant Logic Access

component of the meta-data EIS application model.

The Variant Logic Access uses the following

entities to model the definition of the alternate logic

access:

• User Use Logic Variant: assigns individual

users to access the designated Logic Variant as

their new default for those objects.

• Security Role Use Logic Variant: assigns all

users from the Security Role to access the

designated Logic Variant as their new default

for those objects.

• Security User Account: is the list of Users that

are defined in the application runtime execution

environment.

At runtime, the execution engine identifies

available Variant Logic options for a user and

selects the defined Logic Variant for execution.

3.4 Practical Examples of Variant

Logic in Meta-Data Applications

To illustrate some of the operational benefits of how

users can adjust a meta-data EIS application to more

closely suit their local processes are, in increasing

complexity:

• Application Example: modify the initial

starting Canvas (analogous to a UI screen or

form) of an application.

• Canvas Example: reorder the positioning and

change the sizing of various Freeform Panels

(logical grouping of UI object) on the screen.

• Navigation Panel Example: update the

common Navigation Panel (analogous to menus

or toolbars) with any new objects as new GUI

Reports (analogous to custom reports) and

Canvases.

• Freeform Panel Example: alter the names of

several objects to more relevant local terms as

well as changed the supporting text and help

files. Remove data columns that are not relevant

to local conditions from all Freeform Panels and

from the View Tables (analogous to updateable

database views).

• Freeform Panel Example: modify freeform

entry objects GUI Object (analogous to any UI

object) to become a GUI Selection (analogous

to any selector such as Drop Down Box etc).

• Freeform Panel Example: add new View

Columns (analogous to data table columns) to

record unique information requirements, update

validation functions to help minimise the

occurrence of data entry mis-keying.

• Function Example: define new data columns

an calculation functions to integrate new and

existing data.

• Workflow Example: define new workflows or

insert new workflow stages into existing

workflows.

• Module Example: define a new application

module including; new data objects, all UI

screens, reports and workflows.

By extending the above simplistic examples with

more complexity, including adding entirely new

functionality, the accessibility, power and

immediacy of Variant Logic becomes a key

capability of the meta-data EIS application,

particularly with the empowerment of a much wider

base of new Logic Definers, including

knowledgeable end users.

4 CONCLUSIONS

The Variant Logic capability in combination with

meta-data EIS applications provides a framework

that can allow any level of application customisation

to be securely and flexibly applied to an existing

meta-data EIS application.

While our separate analyses have shown that

meta-data EIS applications can have proportionally

significantly lower lifecycle costs compared to

traditionally developed EIS applications (circa

15%), we believe that the Variant Logic capability

alone will provide additional order of magnitude

tangible efficiency savings due to:

• Providing an open standard of application

definition (for meta-data EIS application

models) that minimises the level of vendor

dependence and increases the availability of

competitive additional application logic

definers.

E25_Logic_Variant

...

E20_Security_User_Account

...

E27_User_use_Logi c_Variant

S02_Application_Security_Role

...

E26_Security_Rol e_use_

Logic_Variant

VARIANT LOGIC META-DATA MANAGEMENT FOR MODEL DRIVEN APPLICATIONS - Allows Unlimited End

User Configuration and Customisation of All Meta-data EIS Application Features

399

• Allowing organisation business analysts and

knowledgeable and power end users to

directly and securely define or modify the

application logic to suit local business

functions, has virtually unlimited potential

to offer increased efficiency due to rapid

concept turnaround into concrete solutions,

and reducing the high cost of using

dependent vendor resources.

• When combined with the ability to merge

multiple meta-data EIS applications, and

their common functions progressively

mapped to a single structure facilitating

rapid data and application integration,

without coding. (Davis, 2005)

• All Variant Logic is also preserved during

automated meta-data application updates

that may be provided by the originating

vendor or developer. Any logical conflicts

that may arise during the update can be

precisely identified to minimise any re-

definition changes that may be required to

the user meta-data customisations to avoid

wholesale re-engineering efforts.

REFERENCES

Dettmer, R., 1995. A Class Act – the Rise of Object-

Oriented Technology. In IEE Review, Nov 1995, Vol

42, Iss 6, pp253-.

OMG, 2010. OMG Model Driven Architecture. In

http://www.omg.org/mda/, 2010.

Mostafa, A., Ismall, M., El-Bolok, H., Saad, E., 2007.

Toward a Formalisation of UML2.0 Metamodel using

Z Specifications. In Proceedings of 8

International

Conference on Software Engineering, Artificial

Intelligence, Networking, and Parallel/Distributed

Computing. Jul-Aug 2007. Vol 1. Pp694-.

Yan, J., Zhang, B., 2009. Support Multi-Version

Applications in SaaS via Progressive Schema

Evolution. In Proceedings of the IEEE 25

International Conference on Data Engineering, Mar-

Apr 2009, pp1717-.

Talevski, A., Chang, E., Dillon, T.S., 2003. Meta model

Driven Framework for the Integration and Extension

of Application Components. In Proceedings of 9

International Workshop on Object-Oriented Real-

Time Dependable Systems.

Rajkovic, P., Jankovic, D., Stankovic, T., Tosic, V., 2010.

Software Tools for rapid Development and

Customisation of Medical Information Systems. In

Proceedings of 12

IEEE International Conference on

e-Health Networking Applications and Services, Jul

2010, pp119-.

Hagen, C., Brouwers, G., 1994. Reducing Software Life-

Cycle Costs by Developing Configurable Software. In

Proceedings of the Aerospace and Electronics

Conference, 1994, pp1182-.

Dittrich, Y., Vaucouleur, S., Giff, S., 2009. ERP

Customisation as Software Engineering: Knowledge

Sharing and Cooperation. In IEEE Software, Nov/Dec

2009, Vol 26, Iss 6, pp41-.

Hui, B., Liaskos, S., Mylopoulos, J., 2003. Requirements

Analysis for Customisable Software: a Goals-Skills-

Preferences Framework. In Proceedings of the 11

IEEE International Requirements Engineering

Conference, Sept 2003, pp117-.

Schmidt, D., 2006. Introduction Model-Driven

Engineering. In IEEE Computer Science, Feb 2006,

Vol 39, No.2, pp25-31.

Ortiz, G., De Prado, A., 2010. Improving device-aware

Web services and their mobile clients through an

aspect-oriented, model-driven approach. In

Information and Software Technology, Oct 2010, Vol

52, Iss 10, pp1080-1093.

Cicchetti, A., Di Ruscio, D., Di Salle, A., 2007. Software

customization in model driven development of web

applications. In Proceedings of the 2007 ACM

symposium on Applied computing, ACM, New York,

NY, USA, pp1025-1030.

Fabra, J., Pena, J., Ruiz-Cortez, A., Ezpeleta, J., 2008.

Enabling the Evolution of Service-Oriented Solutions

Using an UML2 Profile and a Reference Petri Nets

Execution Platform. In Proceedings of the 3

International Conference on Internet and Web

Applications and Services, June 2008, pp198-.

Zhu, X., Wang, S., 2009. Software customization in model

driven development of web applications. In

Proceedings of International Conference on

Management and Service Science, pp.1-4, 20-22 Sept.

2009.

France, R., Rumpe, B., 2007. Model-driven Development

of Complex Software: A Research Roadmap. In

Future of Software Engineering (FOSE 2007). IEEE.

Davis, J., Tierney, A., Chang, E., 2004. Meta-data

framework for EIS specification, In 6th International

Conference on Enterprise Information Systems, Porto,

Portugal, April 2004.

Davis, J., Tierney, A., Chang, E., 2005. A User Adaptable

User Interface Model to Support Ubiquitous User

Access to EIS Style Applications. In Proceedings of

the 28th International Conference on Computer

Software and Applications, Edinburgh, Scotland, July

2005.

Davis, J., Tierney, A., Chang, E., 2005. Merging

Application Models in a MDA Based Runtime

Environment for Enterprise Information Systems. In

Proceedings of the 3rd International IEEE Conference

on Industrial Infomatics, Perth, Australia, August

2005.

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

400