Get flash to fully experience Pearltrees
I was very struck a couple of years ago by a comment made by my colleague Eric Hellman. He talked about metadata in terms of rivers and lakes. In the library cataloging model we have had lakes - accumulating stores of data that do not change frequently over time and are fed by a few principal sources. In the ERM/Knowledge Base model we have rivers - stores of data that change frequently as products and services change and which are fed by many streams.