1 Introduction

"In the beginning the Universe was created. This has made a lot of people very angry and has been widely regarded as a bad move" - Douglas Adams

1.1 A New Electronic Information Age

There is a contemporary style of science fiction writing that has evolved over the past ten years which is known as cyberpunk. This form of writing differs from normal science fiction in two crucial areas. Firstly, it is directly related to the cultural concepts that have developed during the computer revolution of the late 1980s and early 1990s. And secondly, it makes a fundamental distinction about information over technology. In cyberpunk novels, electronic information is the motivating and dominating factor in the world; technology is assumed and is merely an enabling tool.

Cyberpunk writers, such as William Gibson and Neal Stephenson have both used differing terms to describe what is essentially the same concept. In Neuromancer (Gibson, 1984) what Gibson calls Cyberspace and in Snow Crash (Stephenson, 1994) what Stephenson calls the Metaverse, superficially characterise an electronic environment that is accessed through immersive virtual reality.

However, both of these terms have a much deeper meaning and more wide-ranging implications: they are both used to represent the entire sum of human information available in an electronic format. This information is accessible to users through powerful artificial intelligence software that can search and cross-reference information of any format very quickly. Indeed, Stephenson devotes a number of chapters to dialogue between a user and an AI program known as the Librarian daemon. This user has an interactive, real-time, natural English discussion with the Librarian, who collects, correlates and presents information to the user at any level of detail. The information that is available to the Librarian, and hence the user, is historical (through archives), contemporary (through current media input) and real-time (through a global network of satellites and sensors).

Whilst the realisation of such an advanced system is not dealt with at a detailed technological level, it is easy to see how the atomic elements of computer science today (for example, artificial intelligence, information retrieval and high-bandwidth networking) have been interwoven to construct a fictitious, but potentially powerful and useful electronic information environment.

1.2 The Reality of the Current Electronic Information Age

The current electronic information age is going through a radical change. With the increase during the past five years of inter-networked environments, such as the Internet, more electronic information is available than ever before. Computer information, which was primarily text-based until a few years ago, is also expanding into the areas of two- and three-dimensional graphics, sound, video and real-time video. The concept of a global, multimedia information space is starting to take form.

However, computers and networks of information present humanity with a fundamental problem: how is information to be organised, managed, navigated and searched on a global scale? These problems of distributed information management have only ever been considered on small scales previously, for example, libraries, encyclopedias and dictionaries, which for the most part had to be painstakingly compiled and cross-referenced by hand.

As more users join the electronic information age, so the human electronic information space grows. However, the problem is that it is growing in unstructured and unpredictable ways. Therefore, as it grows in an anarchistic manner, so the user finds it increasingly difficult to manage and to navigate and to obtain the information that they require.

Current technology in this area provides relatively few and generally unsophisticated tools to help the user either search for the information that they require, to manage their own sets of distributed information or to integrate with other information resources. Tools that do exist, tend to be aimed at a specific information resource and, currently, very much within the area of information discovery, for example, search engines for the World-Wide Web.

1.3 The Challenge for Distributed Information Management

It is obvious that a mechanism is required that is co-ordinated and developed at a global level that takes into account all current technology, but is flexible enough to be able to also incorporate future advances. However, is this possible or even desirable? How long would such a system take to design, implement and prime? What happens to all of the existing electronic information that currently exists? And, what occurs in the mean time?

Therefore, the immediate and current challenge for a global initiative for managing distributed information must be based upon current and existing technologies, must be available now and must also provide the user with flexible and expandible tools that they can customise to fit their requirements.

1.4 Agents of Change

Much speculative debate has been generated over the world agent. To some an agent is a piece of software that performs a task on behalf of a user. To others it is diverse pieces of software that collaborate to perform a task on behalf of a user. To others still, this piece of software (or pieces of software) must achieve more than simple task completion; they must possess human and mental characteristics, for example, intention, intelligence, autonomy and believability.

The essential difficulty in classifying the role or nature of an agent is difficult because they are generally ascribed tasks that would normally be carried out by a human. It is this quality that makes them difficult to define, but gives them a distinction over other pieces of software. Chapter 3 details the common characteristics that are attributed to agents and discusses their implications.

The introduction of agents represent a fundamental change in the way users will interact with their information; the emphasis will be moved away from a user interactively working with a piece of software to achieve a task to a user delegating the completion of a task to an agent. This implies that agents do have a useful and important role to play within a distributed information management system.

Additionally, agents can overcome boundaries that are difficult for normal software to cross, for example, platform heterogeneity, integration with legacy systems, movement across networks, etc. The ability for agents to adapt themselves to their environment and given situations would make them a powerful and flexible tool for users.

Assuming that the agent-user trust relationship can be developed and preserved, this means that users will benefit from improved management of their distributed information, improved navigation of other distributed information resources and improved retrieval of the actual information that they require.

1.5 Overview

The next chapter of this thesis examines the current distributed information management tools that are available and in common use. It continues by proposing that the concepts of discovery, navigation, consistency and integration are key and crucial aspects that any distributed information management system must posses in order to fulfil the requirements of its users.

Chapter 3 discusses the concept of an agent and how this can be related to a piece of software. Notions of agency are presented that are useful in describing the nature and boundaries of agents to ensure that they can fulfil their tasks in a sensible and predictable manner. The chapter goes on to give a taxonomy of the current perceptions of agents by the differing communities that are using them.

Chapter 4 discusses how a meta-characteristic of agents, namely that of mobility, can be employed to extend the functionality and boundaries of domains within which an agent can work. It details how mobile agents are a direct evolution from the client/server paradigm by moving the processing closer to the resource. The characteristics that mobile agent systems embody are described and the state of the art in mobile agent technologies are then presented and compared against these characteristics.

Chapter 5 describes how an architecture can be developed to support mobile agents and distributed information management. The requirements of such a system are given and then the details of a framework are outlined. This chapter hypothesises that a basic agent infrastructure must exist before integration with distributed resource systems can be accomplished.

Finally, chapter 6 outlines the future directions of this research into a prototype to realise some of the goals identified earlier. Chapter 7 presents a final overview of the research that has been undertaken and presents a summary of the conclusions that have been drawn.

[Top]

[Back]

[Next]

[Bottom]

EMail: jd94r@ecs.soton.ac.uk
WWW: http://www.ecs.soton.ac.uk/~jd94r