Document Sample
References Powered By Docstoc
					Efforts Towards an Open Electronic Commerce Infrastructure for the Tourism
          Penelope Constanta, Manolis Marazakis, Dimitris Papadakis, Christos Nikolaou
                             Institute of Computer Science, FORTH,
                         PO Box 1385, GR - 71110 Heraklion, Greece
                    E-mail: {penelope,maraz,dimpapa,nikolau}

                                           Dietmar Boenke
                          LTU Lufttransport-Unternehmen GmbH & Co. KG.
                           LTU-EDV-Betriebswirtschaftliche Anwendungen
                           Parsevalstrase 7B, D-40468 Dusseldorf, Germany

This paper motivates the development of an open infrastructure to support electronic commerce for the
tourism industry. Such an infrastructure is the objective of the OnTour project, involving representatives of
all participants in the tourist service supply chain. In this paper we discuss unique aspects of service
provisioning in the tourism industry, and demonstrate the need for a business process management
infrastructure to manage customer-tailored requests in the context of highly dynamic networks of service
providers. We also provide a brief outline of Aurora, a related research project aiming to provide a common
infrastructure for network-centric applications.

1. Introduction
In an open market, there are many providers of goods and services, and intermediaries providing more
complex services by combining offerings from multiple providers. Moreover, clients themselves may
combine on-demand offerings to realize composite product/service packages. The Internet provides a unique
potential for electronic commerce, by providing a common environment for a large population of clients and
providers. Appealing aspects of this environment include the fact that many offers are available for clients,
enabling them to search for “bargains”, and, from the providers' point of view, a large number of potential
customers and the possibility to reduce both business-to-customer and business-to-business transaction costs
by establishing largely automated processes. However, it is commonly accepted that the currently available
infrastructure for electronic commerce does not live up to the expectations of customers and providers (for
example see [4]).
Current electronic commerce (EC) systems do not support the notion of a complex product/service
consisting of items from multiple providers. They are limited to commerce transactions over items in a
single catalog or in an electronic mall that hosts multiple stores. Realizing complex commerce transactions
among multiple providers involves a complex sequence of actions, with interdependencies that span over a
long period of time each action involving interactions with information systems and personnel. We envision
an open infrastructure that will enable searching and advertising, negotiating, contracting and ordering,
billing and payment, distribution and receipt, accounting, and other customer services. As in the case of DL
systems [12], for this view to become reality, it is important to acknowledge that performing commerce
transactions is a process that involves multiple participants and phases [14, 15]. This view requires
substantial system support to provide interoperability among distributed autonomous providers for
coordination and negotiation. Current EC infrastructures do not provide the required support for
coordination and negotiation in open environments. They have emphasized interoperability at different
levels of abstraction [6, 1, 7] and electronic payment and delivery protocols [5,9].
In Section 2 we provide an outline of the OnTour project, which was recently granted funding by the
European Union to develop and evaluate, in a full-scale commercial environment, an open commerce
infrastructure for the tourism industry. The tourist market is an area of worldwide importance and growing
turnover, and poses unique challenges for an open infrastructure enabling interoperation among independent
business partners. Companies being active in the tourism market require such an infrastructure for reasons of
competition, satisfaction of the source markets, and optimization of co-operative structures. Section 3
provides an overview of Aurora, a research project currently under way at FORTH/ICS focused on the
design and implementation of an open infrastructure for network-centric applications. The work carried out
in the context of Aurora is highly relevant to the infrastructure requirements for an open market of services
envisioned by the OnTour project. Section 4 concludes the paper by summarizing the ideas expressed and
emphasizes the mutually supporting relationships between the two projects presented.

2 The OnTour Project
The business of travelling is based on the cooperation of commercial organizations building the tourism
value chain. The business partners and their roles are the following:
    •= Suppliers are the source tourist services. Their role is the physical making of the service like
       transportation, car rental, and accommodation.
    •= Tour Operators are the producers of salable tourist products, which are combinations of services
       offered by suppliers. Particular objectives are the customization of services concerning times of
       travel and the optimization of the utilization of bought capacities. It is important to note that in the
       service industry there is no capability of storing “products”.
    •= Travel Agencies are acting as broker tour-operating services and as consultants concerning travel
    •= Travelers are the final customers of tourist services.
The economy of tourist services depends on the ability of companies to adapt their offerings to market
changes on extremely short notice. This means that the capability of bringing up-to-date short-lived
information to the interested parties in the tourism market is an essential requirement for the entire value
chain. Thus, the added value of tourist services is mainly derived from the efficient combination of up-to-
date information that guides the related business processes in this value chain. It is evident that this
requirement necessitates a system infrastructure enabling on-demand combination of information and
services. Due to the autonomy of providers, peer-to-peer interactions among providers are required to
realize the business processes of the value chain.
The Open Network for Tourism (OnTour) project (ESPRIT Project 26956) aims to develop and validate
new ways in the business between suppliers, vendors and customers in the tourist value chain. Therefore, an
open network for electronic commerce in tourist business will be developed, offering new access to and
distribution of information for the partners of the value chain. The network will integrate and facilitate
interoperability of existing technologies to give also new impulses for innovating the workflow of other
companies and branches of business characterized by supply chain management.
Today, the process of selling suppliers' services to customers consists of the following steps. Tour operators
buy capacities from the suppliers. These capacities include transportation to and from the destination (for
example using the services of airlines) and accommodation and miscellaneous arrangements at the
destination (for example rooms in the hotels, bus transfers, rental cars, round trip tours etc). The processes of
buying and organizing services in the destination are supported by the destination agencies that are also
looking after the travelers during their stay in the destination. The bought capacities and services of the
suppliers are combined by the tour-operators. Their products are offered to the customers as a package tour
with a defined category of rooms and a definite time of stay or as a travel package with variation options.
Variation options concern the category of rooms, the time and duration of the stay and round trip tours or
other individual needs. Travel agencies sell the products of the tour-operators and hand over the travel
documents to the traveler. These documents (vouchers) are a confirmation for the right to use a specific
service at the destination. The suppliers and agencies in the destination are informed by the tour-operators on
the services (quality and volume) that have been booked and need to be prepared. For example, hotels get
rooming-lists with information on which guests are to be expected and what categories of rooms are
requested for them.

The OnTour network (see Figure 1) will allow all kinds of companies in the tourist business to offer and
request services, enabling especially smaller and medium-sized organizations to participate in the business
as equal partners. Additionally, the traveler will get the opportunity to communicate directly with destination
agencies and suppliers, such as car rental companies and hotels. Thus, OnTour will offer a fully integrated
solution supporting all communication in the tourist business chain. Entire business processes, such as
providing information to the tourist, booking the journey, services provided at the destination and all the
necessary payments will be carried out via the OnTour network. This functionality provided by OnTour
cannot be substituted by already existing solutions. These only provide subsets of the required functionality,
such as travel offers but do not support booking, services, accounting transactions and electronic payments
in an integrated manner. Existing systems are of proprietary character and can only be used between specific
collaborating companies, which makes it very difficult for other partners to enter this business value chain.

                          Figure 1. The OnTour Business Network.

The OnTour project consortium consists of 7 partners and 4 associated partners from 3 European countries.
Each member is uniquely placed to exploit the results of the OnTour project to the overall extent from
differing yet complementary perspective. The industrial partners LTE (Spain, charter airline), LTU
(Germany, an integrated tourist company), Viajes Necan/VSS (Germany, management consultant company),
Zeus of Crete/ZOC (Greece, destination tour agent) and NE (Spain, destination tour agent) are the main
actors in a highly dynamic travel market and are faced with current lack of integration. The software and
consultancy company VSS brings in their specific expertise in software development. They will be assisted
by BIBA (Germany) and ICS/FORTH (Greece) as research organizations with experiences in the field of
Inter- and Intranet-Technologies. The four associate partners include hotel chains and car rental offices.
Compared with the communication in the business processes of today, the Open Network for Tourism
allows direct communication among all partners of the network. Therefore, it becomes possible to conduct
business transactions not only along the classic connections of business but also along every useful
combination of partners in the business network. The Internet will provide the communication infrastructure
for OnTour, allowing all potential partners to participate with minimal cost and complexity. OnTour will
provide search and selection functionality to assist customers in locating service offers of interest, interfaces
to the underlying commercial IT systems (for reservations, book-keeping, accounting, and payments), and
will demonstrate interoperability in a full-scale business environment by automating business transactions.
By using electronic media instead of paperwork the handling and the functionality of documents is
enhanced. Travelers get a smart card instead of tickets and vouchers. The card can also offer additional
functionality such as credit-card, phone-card, room key, and can be used to automate payments. OnTour
will also address the handling of security problems in public networks. Information concerning booking
services like reservation dates and payment procedures must be protected against misuse. Access to such
confidential information has to be restricted to authorized staff. Therefore, OnTour will integrate
commercially tested security technology in its infrastructure.

3. The Aurora Infrastructure Project
The infrastructure requirements posed by the commercial OnTour project are within the scope of a research
project currently under way at ICS/FORTH. We expect to see a form of synergy between the two
development efforts, with a steady flow of ideas and results passing between them. In open and dynamic
environments, such as the Internet, a network-centric application paradigm emerges. Applications are
created by building new services (either client- or server-side) that transparently integrate and customize
available services. The client then uses the open Internet protocols to interact with the server-based parts of
the application as well as the standard services of the Internet. Applications, such as electronic commerce,
digital libraries, and scientific collaborative work environments, can be realized by combining on-demand
distributed services. This is in sharp contrast with the traditional paradigm of application development based
on fixed specifications of components and their interactions. The developer's task is to identify components
and plug them together, via scripting. This paradigm requires substantial support to identify components at
run-time and dynamically establish communication channels among them, so that they can interact and
coordinate. The execution model has to support peer-to-peer interactions between loosely coupled
participants. This model is in contrast to the more common tightly coupled client/server model. Current open
distributed application frameworks do not adequately support coordination and collaboration, as they assume
well-structured applications where components are developed according to given specifications and interact
in predefined ways. Applications are structured as networks of components with static interconnections and
there is no support for dynamic adaptation of such networks. Figure 2 illustrates the basic components of the
Aurora architecture [8].

                         Figure 2. Overview of Aurora Architecture.
Aurora is a research prototype, currently under development at ICS/FORTH, of an architecture for network-
centric applications. Aurora offers an application paradigm based on the notion of service flow, to realize
composite product/service packages by combining on-demand multiple offerings by autonomous providers.
It supports an open market consisting of multiple autonomous providers of goods and services. Aurora
complements the CORBA and WWW/Java frameworks with support for coordination and collaboration, and
addresses the requirements of dynamic open environments with multiple autonomous service providers. The
technical design for the Aurora architecture is presented in detail in [11, 8]. [13, 10] discuss electronic
commerce in Aurora.
By providing a shared workspace for independent and autonomous participants (customers, providers,
brokers), the service flow paradigm of Aurora addresses the requirements of electronic commerce in an open
environment. Work sessions are driven by events according to specifications expressed using the HERMES
language [8] in the form of Event-Condition-Action (ECA) rules [3]. Events are combinations of service
request messages, state transition signals associated with tasks, as well as system-generated or application-
specific notifications. Conditions for executing actions are expressed as Boolean expressions over events.
Actions activate components, by invoking methods through their exported service interfaces. Thus, ECA
rules drive the flow of control and data between components, in the context of a work session realizing a
composite service (such as a commerce transaction).
The actual actions to be executed and the flow of control and data are determined only at run-time, via the
execution of rules that specify control and data flow. Thus, the course of a workflow may change
dynamically, as the ``next'' actions to execute are determined as the result of rule triggering. Furthermore,
the results of the actions of humans as well as automated tasks may affect the course of a workflow. By
allowing the modification of components and their interconnections and scheduling rules at run-time,
applications become more adaptive to the dynamics of the execution environment. The run-time
environment of Aurora assumes the responsibility to enforce the terms and conditions specified in service
interfaces, gather and manage event data, interpret rules, and trigger the execution of actions for the
A federated directory service (based on the CORBA Trader Service [2]) enables service providers to publish
their services and clients to search for offers of interest. A service provider can assert values for properties of
the service it is advertising. By querying the directory, a client can obtain these values about a service and
constrain its search for appropriate offers based on the property values associated with such offers.
Moreover, a client is informed about new offers or new services published by multiple providers. In the case
of electronic commerce, providers are merchants advertising their products and their services, while clients
are customers searching for certain products or for potential bargains. Service properties are thus prices and
attributes describing the products and services for sale.
A binding service facilitates clients to find service offers of potential interest. Clients can provide the
properties and the desirable values that best describe a service of interest; the binding service then selects
among candidate service offers by matching the properties and the values of clients' requests with published
offers. Clients can withdraw and modify requests that have not been serviced. Moreover, they are informed
about the progress of their requests. The Aurora directory and binding service provide the means to realize
an open market paradigm. Merchants and customers are provided with the means to advertise, search, and
bargain, that are essential to trade in an open market. Furthermore, Aurora is an open architecture that
enables integration with standard commerce services like payment.

4. Conclusions
The digitalization of economic transactions has far-reaching effects on the tourist market. Electronic
commerce will not allow some classic business to act like before but will bring a lot of opportunities to
companies and travelers. The concentration on special competencies and the widening of the distribution of
tourist services will bring a globalization of resourcing markets and distribution markets. In this sense the
OnTour project will not only help to design one part of the electronic commerce in tourism but also can be
a basis of experience for other branches.
The infrastructure requirements of OnTour are within the scope of the Aurora infrastructure project
currently underway at FORTH/ICS. Aurora aims for a unified treatment of network-centric applications, by
providing services to enable dynamic composition and reconfiguration, as well as run-time monitoring, of
work sessions. A steady flow of ideas and results between the two projects will certainly be of benefit to
both projects.

[1]   E.S. Chung and D. Dardailler. “White Paper : Joint Electronic Payment Initiative (JEPI)”, 1997.
      Available via URL
[2]   “Trading Object Service Specification”. In CORBA services: Common Object Services Specification.
      Object Management Group, Framingham, Mass., 1996
[3]   U. Dayal, M. Hsu, and R. Ladin. “Organizing Long-Running Activities with Triggers and
      Transactions”. In Proceedings of ACM SIGMOD Conference on Management of data, 1993
[4]   H. Garcia-Molina, S.P. Ketchpel, and N. Shivakumar. “Safeguarding and Charging for Information on
      the Internet”. In Proc. IEEE Int'l Conf. Database Engineering, (ICDE), 1998. Available via URL
[5]   P.J.M. Havinga, G.J.M. Smit, and A. Helme. “Survey of Electronic Payment Methods and Systems”.
      In Proc. Euromedia Conference, 1996
[6]   CommerceNet Inc. “eCo System: CommerceNet's Architectural Framework for Internet Commerce”,
      1997. Available via URL
[7]   S.P. Ketchpel, H. Garcia-Molina, A. Paepcke, S. Hassan, and S. Cousins. “U-PAI: A Universal
      Payment Application Interface”. In Proc. 2nd USENIX workshop on Electronic Commerce, 1996
[8]   M. Marazakis, D. Papadakis and C. Nikolaou. “The Aurora Architecture for Developing Network-
      Centric Applications by Dynamic Composition of Services”. Technical Report TR 213, FORTH/ICS,
[9]   W3C Electronic Payments Area, 1997. Available via URL
[10] C. Nikolaou, M. Marazakis, and D. Papadakis. “Synthesis of Services in Open Commerce
     Environments”. In Proc. Special Track on Electronic Commerce on the Internet, EUROMED-NET'98,
     Nicosia, Cyprus, 1998.
[11] C. Nikolaou, M. Marazakis, D. Papadakis, Y. Yeorgiannakis, and J. Sairamesh. “Towards a Common
     Infrastructure to Support Large-Scale Distributed Applications”. In Proc. European Conference on
     Research and Advanced Technology for Digital Libraries, 1997
[12] A. Paepcke. “Digital Libraries: Searching is not Enough”. D-Lib Magazine, May 1996. Available via
[13] D. Papadakis, M. Marazakis, C. Nikolaou, and P. Constanta. “An Infrastructure to Support
     Collaboration and Co-operation in Electronic Commerce”. In Proc. Working Conference on Electronic
     Commerce, Metsovo, Greece, 1997
[14] D. Schutzer. “A Need for a Common Infrastructure: Digital Libraries and Electronic Commerce”. D-
     Lib Magazine, April 1996
[15] The NII Cross-Industry Working Team. “Electronic Commerce in the NII”, 1995. Available via URL