Dissertation

Topic

A Model for Adaptive Application on the Semantic Web

Abstract

The PhD thesis aims at introducing a model that gives an abstraction to the functionalities and data involved in adaptive applications for the Semantic Web.

Introduction

The World Wide Web is (likely to stay) a network richly provided with heterogeneous and distributed informational and increasingly, transactional resources (services), which bring together enterprises, people as well as software agents: there is a steady rise in B2B transactions, enabled by the across-company exchange of data over a mutually agreed format and recently, the story even goes further, when a few companies make use of the internet platform to move forward towards a across-company integration of their entire information system so as to achieve strategic alliances. Also, the acceptance of e-commerce increase at the end customer site, as indicated by the increase in the transactional volume in the B2C segment. With the advent of the enterprise portal, the entire information infrastructure of the company is also moving toward the web platform, offering a unified and standardized access for customers, suppliers and employees to the information and services the companies offer. Making use of applications offered over the employees portal, for instance groupware such, staffs members can meet, cooperate, and coordinate their work over the internet. Finally, there are platforms on the internet which are designed to support the cooperative work of software agents. By the development meta standards, more and more meaning can be attached to the information content, i.e. more and more machine understandable resources will become available on the web. Thus , it is likely that agents will behave even more intelligently and autonomously, and perform even more tasks for the human user automatically.

Recent development within the semantic web community suggest that internet, the WWW, as it is today, is about to change. The content and services offered over the internet will be annotated with description , with metadata, which can define the structure, the meaning, the layout and in the case of services, the functionalities and other information related to the resources. This raises questions about the structure of such a Semantic Web.

Traditionally, the web is seen as a collection of linked nodes, as entailed by the specification of the reference model for hypertext applications, the Dexter Model. Most applications on the internet correspond to this model, i.e. offer functionalities and data specified in this model. They deliver a set of documents (nodes) that are linked with each other as well as with other documents outside the domain, over the internet to the user interface, which is the browser.

The web has been a “viable platform” and thus, this model has been successful in delivering resources to the web, for human users to browse.

However, the supply of information is steadily increasing, but so does also the automatic information processing capabilities. According to Berners-Lee, “The Semantic Web is a giant distributed databases that deliver resources that can be understood and reasoned about by machines”. So, there is and will be great potentials to make use of these automation capabilities in order to extract from the overflow of the web the information and services relevant to the user on an ad-hoc basis, and deliver them over a standardized user interface. The Semantic Web software agents to interpret resources in many ways, and adaptation can make them personalize the interpretation to the user’s needs. This suggests that the Dexter Model can no longer abstract the functionalities and data required for this kind of adaptive applications for the Semantic Web. Instead, a new model that extends the Dexter model with adaptive functionalities such as and additionally , incorporates technologies of the Semantic Web so as to define its static and dynamic structures, can guide developers in designing applications that offer adaptive delivery of informational and transactional resources to the internet user.