Introduction

SWAP combines two highly successful technologies, viz. Semantic Web and Peer-to-peer computing. SWAP will develop the technology that is necessary to allow users their individual views on knowledge AND let them share knowledge effectively. In order to allow for individual views, every user抯 PC is treated as a peer and every user may ask queries to the network of peers. In order to actually find the right piece of knowledge, SWAP employs Semantic Web technology. Current peer-to-peer technology only allows for keyword search and lacks the semantics that is necessary for effective, precise knowledge sharing. Current knowledge repositories lack the capability to provide really individual views in a decentralized framework with low administration overhead. SWAP will produce both.

The combination of Semantic Web and Peer-to-peer is highly innovative with prospective benefits to the individualization of work views as well as to the facilitation of knowledge sharing between peers. SWAP will tackle the semantic challenges brought up by this novel combination such that knowledge finding and sharing is effectively possible. The basic infrastructure will build on already available Open Source peer-to-peer solutions. The work package structure of SWAP has been constructed in order to handle the high degree of innovation set out in the project goals, viz. it allows for fast adoption of new techniques and scientific achievements, but ensures effective development of the envisioned objectives. In particular: 1. SWAP will offer a Conceptual Analysis of the problem. The analysis comprises the use of semantic web technology in peer-to-peer computing, but it will also consider how emergent semantics may be constructed by overlapping knowledge structures of different peers (e.g. common categorizations of knowledge, common bookmark names of web pages). 2. SWAP will explore the conceptual architectures made possible by this combination. They constitute the foundations for the subsequent integration task as well as for further exploitation. 3. SWAP will investigate new methods to realize a Semantic Web based peer-to-peer system. 4. The methods will be implemented in a set of tools. 5. The tools will be integrated based on the foundations from the conceptual architectures in a demonstrator that can be tested in case studies. 6. The tool environment will be evaluated in two case studies about Knowledge Management for investment analysts and about Knowledge Sharing between SMEs (virtual enterprises).

Concrete results of SWAP will be:

A comprehensive study of the potential of Semantic Web and P2P for KM and a corresponding methodology.
Method descriptions and software prototypes for P-Oc, P-Al, P-Map, P-Viz, P-I, P-Up and P-S.
An integrated software environment for semantic P2P, rigorously tested and validated in real world KM scenarios.
Objectives

In today抯 knowledge-based economy, the competitiveness of enterprises and the quality of work life are directly tied to the ability to effectively create and share knowledge both within and across organizations. Emerging peer-to-peer solutions are particularly well suited to the increasingly decentralized nature of today抯 organizations, be it a single enterprise or a dynamic network of organizations. They make it possible for different participants (organizations, individuals, or departments within an organization) to maintain different views of the world while exchanging information. They also circumvent the bottlenecks associated with more traditional solutions, which rely on one or a small number of centralized servers. At the same time, because they rely on keyword search and rather simple knowledge representation techniques, today抯 peer-to-peer solutions are extremely limited. They cannot easily support the introduction of new concepts, make it difficult to determine whether two terms are equivalent, and generally can only support very limited levels of automation ?all types of functionality, which Semantic Web technologies have been shown to support.

SWAP is about demonstrating that the power of Peer-to-Peer computing and the Semantic Web can actually be combined to support decentralized environments where participants can maintain individual views of the world, while sharing knowledge in ways such that administration efforts are low, but knowledge sharing and finding is easy. Key to the success of combining Peer-to-Peer solutions with Semantic Web technologies is the use of Emergent Semantics. Emergent semantics builds on light-weight and/or heavy-weight ontologies that different individuals, departments, or organizations have created. It considers the overlap between ontology definitions and the use of concepts and relations with actual data in order to extract shared ontologies for sets of individuals or groups of people. Intelligent tools will use such definitions to ensure that knowledge will be appropriately structured, so that it can be easily refound. Knowledge Management can occur in a distributed fashion without overhead through central administration.

Two case studies of knowledge intensive work, knowledge management for investment analysts (called KM case study) and knowledge sharing between SMEs (called virtual enterprise case study) will evaluate and improve the new combined technology