The Associator Project Home Page

Status: Draft
Version: 1.0, Feb 03
Author: Leo Meyer, meyer@software-services.de

What?

The Associator is a Knowledge Management System aiming at creating a comprehensive database of world knowledge pertaining to all subjects, all languages and all scripts. Main features of this system are:

• It can be used likewise as personal or corporate knowledge base for arbitrary content
• It integrates smoothly with existing knowledge resources
• Every bit of information (words, phrases, texts, multimedia content) is stored along with its context
• Precise context information (definitions, explanations, further information etc.) is available via hyperlink on every bit of information
• Automated translation of text will be much more exact because word context is known
• Semi-automatic generation of context using heuristic algorithms is possible
• Data is highly structured and can be traversed by agents for data mining purposes
• De-centralized servers provide a scalable web of knowledge
• Anybody can publish information and receives payment for it when it is accessed
• Anybody can easily access and download published  information worldwide

Why?

"Dr. Raghavan - Verity's Vice President and CTO [...] will present, how enterprises can attain a safe and efficient Knowledge Management System by application of three main keys. These are according to Forrester Research: Precise searching and finding, organizing content in hierarchical taxonomies and combining human abilities and content in a Knowledge Management System."
- from a press review of a WKF (World Knowledge Forum) article. My translation and emphasis.

The internet (or more specifically, the World Wide Web) has certain disadvantages in distributing and storing knowledge. Some of these are:

• Hyperlinks depend on the mercy and/or knowledge of the page author
• Hyperlinks tend to be broken as pages move, user accounts are deleted or servers are restructured - information is lost
• Available information is redundant
• There is little quality control regarding content of published web pages
• Information, when out of date, hardly contains references to newer resources
• Searching is clumsy and tedious because it is not context sensitive
• There is no or little reward for digitalizing and publishing existing analog knowledge resources, these relying on funding rather than on consumer-oriented market mechanisms, keeping important knowledge out
• Organizations tend to keep proprietary information out of the web, because they derive no benefit from its publication
• Payment for information is insecure and decentralized, thereby creating much overhead
• Interaction with web content is a one-way road with little feedback
• The WWW evolves into a business platform, essentially shifting towards other purposes than knowledge retrieval

For all these reasons the World Wide Web in its current form, consisting of many millions of individual contributors, provides no reliable and effective store for knowledge. Though the amount of data increases daily, its structure does hardly improve, leaving us with terabytes of unconnected data which it is very difficult to scan for what we really need at a given moment.

It should therefore be the aim of everybody who is interested in preserving, extending and distributing the treasure of human knowledge to create an information infrastructure which avoids the above disadvantages. This infrastructure is not intended to make the WWW superfluous, nor to exploit its available resources, but to extend its capabilities by providing a common standard for knowledge exchange.

I am proposing a draft of such a system on these pages.

How?

First we have to decide what we actually regard as "knowledge" and what not. For this purpose, some theoretical ruminations are necessary: Data, Knowledge and Information.

Integrating a Knowledge Management System, Information Retrieval System and Artificial Intelligence

Some notions about Artificial Intelligence and how it may be implemented in the Associator.

Project History