My ideas about what the contents of a global semantic network might look like go back to the late eighties and early nineties. At that time it became already clear that traditional relational databases would probably not be able to fulfill the total information need, in particular with a multi-lingual, multi-domain and historical point of view. Although relational databases were and are perfect for restricted domains and a particular use, beyond that limited scope, they create more problems than that they solve, at least in the way they are generally used. Since that time the practical use of Notion System (the precursor of the current semantic network) showed some clear do's and don'ts and helped to forge ideas about what seems a good practice for the way to store information in a semantic network.
The basic concepts for the information architecture were developed in 1988 and 1989. The next year I started building a system in HyperCard on a Macintosh S30 (Apple). In 1991 this software called Notion System was up and running. Over the next few years the system grew in possibilities and the network it maintained in size. By the end of 1996 the network consisted of about 200.000 nodes and some 500.000 relationships.
From 1997 to 2001 the activities around Notion System were reduced due to other occupations. Finally in 2001, at TNO, we took up the challenge to rebuild a similar system. This time in Java in order to be platform independent and with a non proprietary interface: a web browser. Since that time the system is ever growing in functionalities and the network in content.