AN E-BUSINESS FRAMEWORK DESIGN USING ENHANCED

WEB 2.0 TECHNOLOGY

Karim Ouazzane, Muhammad Sajid Afzal, Hassan Kazemian

London Metropolitan University, 166-220 Holloway Road, London, U.K.

Jun Li

Department of Land Economy, University of Cambridge, 19 Silver Street, Cambridge, U.K.

Keywords: Request brokering, Service adaptors, Workflow management, Customized business logic, e-business.

Abstract: The main aim of this research is to develop a robust, reliable, efficient and novel framework by using Web

2.0 technology that will serve as a front and middleware collaboration model between data persistence logic

and operational requests. This framework will serve as a mediation platform for request brokers. It will

provide a high level of abstraction by encapsulating low level details of the system, such as request

handling, request mediation, response handling and service loading. In order to overcome the hard coded

service mapping with interface, there are no customizable business logic and no generic customized

workflows problems. These are the most essential requirements of converting the Small and Medium

Enterprises into one e-business platform swiftly.

1 INTRODUCTION

In the current era of state-of-the-art cutting edge

technologies, business organizations are rushing to

transform their companies into e-business. Presence

of e-business brightens their chances of winning

market shares by maximizing product availability

and providing related services at convenient and

cheap way to their customers.

However in this race, the e-business growth

pendulum is only swinging one way. This is very

easy to understand by observing today’s business

market. Due to a limited budget, small business

organizations (e.g. local retail shops) cannot afford

costly IT systems and associated

maintenance/administration costs. But despite these

financial constraints, they desperately need

computation to survive and to compete with giant

competitors by expanding their businesses.

Web 2.0 is a second generation design patterns and

business model for web applications. The terms first

coined by (O'Reilly, 2005). Every technology

requires a clear and effective model for representing

its components and the interaction between these

components. Until recently, there has been no

innovative Web 2.0 model which performs

• On-demand customization of business logic

and user interfaces.

• Request management component.

• Dynamic operational request mapping with

the interface.

• Encapsulation of standard business logic.

Unfortunately not much has been done by the

research community in building a clear

understanding of Web 2.0 and its application

paradigm. Every technology requires a clear and

effective model for representing its components and

the interaction between these components. Until

recently, there exists no such proper Web 2.0

framework with the features described above. As a

result, a research conducted by (Omar, Abbas and

Bendiab, 2007) proposed the model framework for

defining Web 2.0 components and their relations as

illustrated in Figure 1.

One of the major deficiency of the Web 2.0 model

framework is none presence of request management

mechanism. According to a research conducted by

(Alur, Curpi and Malks , 2001) , request management

Ouazzane K., Afzal M., Kazemian H. and Li J..

AN E-BUSINESS FRAMEWORK DESIGN USING ENHANCED WEB 2.0 TECHNOLOGY.

DOI: 10.5220/0003417800970102

In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 97-102

ISBN: 978-989-8425-56-0

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

Figure 1: Model Framework by Omar, Abbas, & Bendiab

(Omar, Abbas and Bendiab, 2007).

in web-based systems is a two step activity - Action

Management and View Management.

There are only two types of request management

techniques that have been discovered so far. One is

called ‘conventional’ and another is called ‘request

brokerage technique’.

Conventional request management technique is a

technique that is commonly used in traditional web-

based systems. In a conventional technique, request

management is automatically performed by the

application server.

This concept of request management mechanism on

the basis of request broker’s based system

architecture becomes the centre of attraction for

relevant research communities and still gaining

popularity. Request brokerage in service oriented

architectures is an active research area these days,

such as in ACTS/ABS (Architecture for information

brokerage service) (ACTS, 2000) and have managed

to propose some stable models in which a request

can be dispatched to a best suited service(s),

transparent to a user, with the help of request

brokers.

Web Services are a new emerging web programming

paradigm based on the concept of Service-Oriented

Architecture (SOA) (Bell and Michael, 2010).

(Howard and Kerschberg, 2004) proposed a

complete framework called Knowledge-based

Dynamic Semantic Web Services (KDSWS).

Another research attempted by (Beck, et al, 2000)

considers the idea of next generation electronic

brokerage for performing active and real-time

functionalities. Request broker model proposed by

(XiaoQin, LinPeng, and Minglu , 2004) is a concept

of agent based web-services platform request

brokers model which is being used in this research

and is actually an extension of architecture for

information brokerage service models proposed in

ACTS/ABS.

A research conducted by (Zhao and Tong, 2007) has

proposed a service composition model called ‘A

Dynamic Service Composition Model Based on

Constraints’. This model is not a request broker

based model, but is capable of handling complex

situations. Problems of Web2.0 framework without

request broker are illustrated in Figure 2.

Figure 2: Application model framework without request

broker.

A review of the current Web 2.0 retail systems that

are built upon existing SOAW2 model framework

has shown the lack of fundamental functionalities

such as on-demand customization of business logic

and user interfaces. A system like ZohoCRM has

shown some capabilities of user interface

customization but it is only limited to the

customization of business reports (A.H.M. van

Limburg, 2010). Furthermore, another important

observation is that none of the current Web2.0 retail

systems is capable of providing on-demand

customization of business logic. This is because of

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

the fact that applied SOAW2 model framework does

not support explicit request management. Current

SOAW2 model framework directly exposes user

interfaces to business services that reside inside

resource container. Also, it is crucial to note that the

presence of supporting function between user

interfaces and resource container should be ignored

as they don’t perform any special tasks with the

exception of searching & binding of best suited

services. This direct exposure led to a result in the

form of user interfaces that contain concrete

business services mapping instructions. Due to this

fundamental shortcoming of the framework,

developed systems do not provide on-demand

customization of business logic at run-time.

Moreover the framework does not provide any extra

layer to secure this customized business logic.

Therefore, an investigation is required to propose a

new Web 2.0 framework that will overcome the

current SOAW2 fundamental shortcomings (D.

Gallula, 2009; K. Ducatel, et al, 2010).

2 THE PROPOSED

FRAMEWORK

2.1 Fundamental Architecture

The proposed model framework is based on SOAW2

and contains fundamental changes in existing

models framework. The fundamental changes

include introduction of effective and intelligent

request broker architecture and the replacement of

supporting functions with service adapters, as shown

in Figure 3.

Figure 3: The proposed model framework.

By taking into consideration the shortcomings of

request management component in Web2.0

framework and a need of effective and intelligent

request brokerage mechanism to handle complex on-

demand sharing and customization problem, a new

architecture of the framework is proposed which

overcomes the inadequacies of the existing Web2.0

framework. The proposed model framework

integrates request broker architecture in between

user interface and resource container. This will

increase the action and view management of the

model framework. It might also provide on-demand

request routing between user interfaces and services

of standard platform. The second important

component in this proposed model framework is a

replacement of supporting functions with service

adapters. This component could avoid any conflict

with SOA principles. Also, in this architecture

workflows are modeled as service adapters.

The supporting functions in the new model

framework are replaced with service adapters.

Service adapter is a new concept of light weight

services and contains implementation of workflows.

In SOA tradition, services are relatively large,

shared, intrinsically loosely coupled units of

functionalities, and have no embedded calls to each

other.

It is debateable that proposed model framework

requires sharing of generic workflows among

different companies (i.e. users) then why can’t

generic workflow be modelled as web services. The

simple answer of this question is that, in proposed

framework, workflows are actually sequential calls

to the services of core platform. Therefore, if

workflows get modelled as web services, it will be a

violation of loosely coupled and no embedded calls

principle of SOA.

To overcome this limitation and to avoid any

conflict with SOA principles, all Workflows

including generic and customized are modeled as

service adapters. On a user processing request, these

adapters are connected with the core platform to

execute the modeled workflows.

2.2 A Layered Representation of the

Framework

Layered representation of the framework shown

above is divided into four layers namely

presentation, request management, operational and

core service layers.

AN E-BUSINESS FRAMEWORK DESIGN USING ENHANCED WEB 2.0 TECHNOLOGY

Figure 4: Framework Layered Representation.

2.2.1 Presentation Layer

Presentation layer comes at the top and it consists of

a user interface component that is internally divided

into two sub components – generic user interfaces

and customized user interfaces; Generic user

interface component holds the generic set of user

interfaces. These user interfaces shared across the

interested companies. Customized user interface

component holds the set of user interfaces that are

customized for some companies. Presentation layer

of the system is directly exposed to users thus serves

as a gateway to the system. The users use it for

sending data processing requests to the system.

2.2.2 Request Management Layer

Request management layer consists of request

brokers and UI container components. Presence of

the system manager component on this layer

indicates a control and administration. Request

broker component on this layer is responsible for

brokering user’s requests (received via presentation

layer) to service adapters and then brokering back

the responses to the users via presentation layer. UI

container serves as a data storage that holds the

response data.

2.2.3 Operational Layer

Operational layer consists of profile factory, service

adapters and authentication components. Profile

factory component is responsible for holding the

profiles of the client companies. Authentication

component is part of a system security and provides

assistance to the system manager in authenticating

and authorizing users and their respective locations.

2.2.4 Core Service Layer

This layer is core services layer and it consists of

core platform components. It provides the access of

core business and persistence services to the

components that exist within operational layer.

Moreover, it assists operational layer components in

the accomplishment of their required functionalities.

For example, on one end, it facilitates service

adapter in making business services related calls for

data storage and retrieval, while on the other end it

assists authentication component in validating the

user credentials such as user ID, password & roles.

In addition, it also provides service to the profile

factory component for retrieval of company related

information from database such as location details,

employee details, addresses and contacts numbers.

3 HYPOTHETICAL MODEL IN

ACTION

Any user request for data processing can be divided

into two sections, data part and action part. Data part

contains the data that requires processing. Whereas

the action part describes the required operation on a

given data. On arrival, a user request gets queued up

in a waiting area. The system manager consistently

checks the waiting area and as soon as a request

arrives, it will be allocated to one request broker.

This allocated request broker moves the request

from waiting area into processing area and starts

analysis of a request header to find out the source

details, such as, the name of user interface from

which the request is being generated and the action it

requires. On identification of source and action, it

starts searching to find out the name of matching

service adapter in a user session profile. User session

profile is a profile carried by each user and it is

initially allocated to them by the system manager

when they first log in.

On successful match, request broker binds the

relevant service adapter (i.e. either generic or

customized) and executes it by providing data and

action part of the request. This binding of service

adapter is called ‘request brokering’. The service

adapter only requires data and action to be executed.

Hence, they are services that are independent of their

usage scenario and can be used by any user. This is

one of the main feature of the proposed Web2.0

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

100

Architecture for Service and View Brokerage

(W2ASVB) model framework and it promotes the

idea of multi-company sharing service adapters with

the exception of customized ones as they are

company-specific implementations and hence

confidential. Figure 5 illustrates the flow of control

in request brokering mechanism.

Figure 5: Request brokering in action.

Figure 6: Response brokering in action.

On completion of execution of bonded service

adapter, request broker unbinds the service adapter

and loads the output data (if any) which are being

generated as a response of the execution of response

object. It then starts searching again the user session

profile to find the destination user interface address.

On successful match, it dispatches the response data

back to the user along with output data. This

unbinding of service adapter and mediation of

response back to the user is called ‘response

brokering’, as shown in Figure 6. Since the user

interfaces that generates request and receives

response data are independent of company usage

scenario, they can therefore be used by any user.

This is another major feature of the proposed model

framework and it promotes the novel idea of multi-

companies sharing user interfaces. Figure 6

illustrates the flow of control in response brokering

mechanism. At the end of response brokering,

request broker releases all the holding resources and

makes itself available to the system manager to be

allocated to another request.

Figures 5 and 6 illustrate four examples of request

mappings. In the figures it is clear that user 1 of

company 1 uses generic user interfaces only.

Whereas user 2 of company 2 uses some generic and

some customized user interfaces; finally user 3 of

company 3 uses customized user interfaces only.

Figure 5 depicts mapping of request R4 (i.e.

generated by generic user interfaces) to generic

service adapter and request R2 (i.e. generated by

customized user interface) to customized service

adapter. On the other hand, Figure 6 depicts changes

in the scenario where request R5 (i.e. generated by

customized user interface) is mapped to generic

service adapter and request R6 (i.e. generated by

generic user interface) is mapped to customized

service adapter. However, the mapping shown in

the examples above are not the boundaries of the

framework, it is also capable of supporting other

possible mappings which reflects its universality.

4 CONCLUSIONS

The investigation started with a question of building

up a model framework by using Web 2.0

technology, which enables small retailer to achieve

e-business transformation. The model framework

provides them with an option to customize their e-

businesses according to their individual needs. With

reference to customization and intelligent request

management, it can be concluded that the designed

system with the application of the new proposed

model framework has proven its capabilities of

handling such complex situations.

AN E-BUSINESS FRAMEWORK DESIGN USING ENHANCED WEB 2.0 TECHNOLOGY

101

Initial version of proposed framework architecture is

presented along with the problem description; it is

evident that the framework requires an effective

model that serves as a front & middleware

collaboration. This model framework provides an

intelligent feature, that is interfacing for sharing core

business logic among all small businesses as well as

providing customization facility without affecting

other businesses.

For future work it is worthwhile exploring the need

of demand-and-supply algorithm to control the

request brokers’ pool size at run-time. Given the

highly operational and business-centric nature of the

system, the authors recommend the use of current

data mining algorithms to achieve an improved

version that not only performs the run-time

statistical calculation of incoming requests but also

uses its own knowledge base to decide on the pool

size.

REFERENCES

ACTS Project (2000). Project ABS (Architecture for

information Brokerage Service) Final Report.

Avaliable:ftp://ftp.cordis.europa.eu/pub/infowin/docs/f

r-206.pdf.

A.H.M. van Limburg, (2010), Towards innovative

business modeling for sustainable eHealth

applications, Second International Conference on

eHealth, Telemedicine, and Social Medicine

Alur, D., Curpi, J., & Malks, D (2001). Core J2EE

Patterns: Best Practices and Design Strategies. USA:

Sun Microsoft Press. 60-65 .

Beck, et al (2000). Active and Real-Time Functionalities

for Electronic Brokerage Design, International

Workshop on Advance Issues of E-Commerce and

Web-Based Information Systems ,

Bell and Michael (2010). SOA Modeling Patterns for

Service-Oriented Discovery and Analysis. UK: Wiley

& Sons. 390

D. Gallula (2009), Enriching the E-learning experience in

the framework of Web 2.0 using Usability 2.0 (in

Hebrew), Dept. of Information Sciences, Bar-Ilan

University, Ramat-Gan, Israel.

Howard, R., & Kerschberg, L. (2004). A Knowledge-based

Framework for Dynamic Semantic Web Services

Brokering and Management. Database and Expert

Systems Applications, 15th International Workshop on

(DEXA'04), pp. 174-178.

K. Ducatel, M. Bogdanowicz, F. Scapolo, J. Leijten, and

J.C. Burgelman (2010), Scenarios for Ambient

Intelligence in

2010,Available:ftp://ftp.cordis.lu/pub/ist/docs/istagsce

narios2010.pdf. Last accessed 10th Jan 2011

O'Reilly, T. (2005). What is Web 2.0? . Available:

http://www.oreillynet.com/pub/a/oreilly/tim/news/200

5/09/30/what-is-web-20.html. Last accessed 4th Dec

2010

Omar, W., Abbas, A., & Bendiab, T. . (2007). SOAW2 for

Managing the Web 2.0 Framework. IT Professional -

IEEE,. 1 (1), 30-35 .

Rezaiian Fardoii, S., Monfared, M.A.S. (2008), Design of

an Intelligent Customer Relationship System Using

Evidential Reasoning Approach, Int. Conf ICDM ,

IMT Ghaziabad, India,

XiaoQin, Lin Peng and Minglu, (2004) , Design and

Implementation of an Agent-Based Web Services

Platform for Electronic Commerce. Services

Computing, IEEE International Conference on

(SCC'04), pp. 643-646

Zhao, H., and Tong, H. (2007). A Dynamic Service

Composition Model Based on Constraints. Sixth

International Conference on Grid and Cooperative

Computing (GCC 2007) , pp. 659-662 .

ICEIS 2011 - 13th International Conference on Enterprise Information Systems

102