Collective Information

A JMC2 Suprema Project, 2004.

Introduction

Collectively a population contains vast amounts of knowledge and modern communications technologies increase the ease of communication. However, it isn't possible for a single person to aggregate the knowledge of thousands or millions and abstract useful information from it. Collective information systems are attempts to harness the knowledge of a population and to present it in a simple, fair and attack resistant manner.

Collective information systems have developed in a number of areas, with a number of different requirements. They can be roughly grouped into:

• Collaborative filtering
• Search engine ranking (PageRank)
• Trust and Reputation

Centralised Systems

Centralised systems have the advantage that information flows to a central location which can then do processing based on the most complete set of data.

Decentralised Systems

Although centralised systems have a number of advantages they do present a single point of failure and a single point of trust. If you are relying on eBay to gauge the trustworthiness of a member then you first have to trust eBay!

In some applications this is unacceptable and decentralised systems have been developed.

References