The Ontolingua Project at Stanford KSL has supported a broad based Web-mediated knowledge acquisition effort [Fikes et al. 1991] [Gruber 1992]. The Ontolingua library consists of several hundred composable "ontology fragments" which may be composed to form core domain theories for a variety of applications. The underlying semantics of Ontolingua are those of a full first-order theory (KIF) but the notation hides this most of the time since part of the goal has been to capture portable ontologies which can be easily exported into a variety of systems. The ontolingua notation is relatively formal as would be appropriate for core domain theories, particularly in technical domains.
Although the Web interface has allowed many people to contribute, it doesn't appear to be the case that much work has gone into management of consistency or tools for developing consensus where possible. The intent seems to be to support a collaborative among knowledge engineers with the view that the assets developed will then be of great use as a bootstrap in building a variety of Knowledge Based systems in a variety of domains.
Our Web based collaboration in contrast will mainly focus on collecting annotation information for a very large collection of information sources. We focus on the use of natural language because we believe that this is a key enabler for a broad based collaboration.