Talk:Rete algorithm

From Wikipedia, the free encyclopedia

You have new messages (last change).

We need a paper on Charles Forgy and also we should probably mention the (commercially available) Rete II algorithm. I wish my explanation of Rete were better, maybe someone else wants to take a stab at it. Chadloder 18:12 Jan 22, 2003 (UTC)

[edit] Changes

I have extensively re-written my original Rete content, in part to provide a fuller explanation, and in part to address some of the feedback I have receieved. For example, the original description of tokens was poor, and I have provided a much more complete description of their use, and different common implementations. I've broken the text up into more digestable chunks, buit I'm not sure it will be any easier to read. Rete exhibits a degree of complexity, and there are several commonom implementation variations, so I'm not sure it is possible to provide an easy read. What would help is a set of images. I intend to change the single diagram and remove the 'dummy' input at the top of the beta network. This lends little to the general understanding of Rete. If I get time, i will create additional diagrams to illustrate the text.

In addition, I plan at some point to add some links to other Wikipedia pages, and also to add a paragraph on TREAT and another on LEAPS.

Charlesyoung 18:09, 27 October 2006 (UTC)

  • I just changed the first line to say that Rete is used for production rule systems instead of expert systems. This seemed right as expert systems can be implemented using any language, whereas Rete really is just for production rule systems. --Ferris37 21:01, 30 November 2006 (UTC)
Retrieved from "http://en.wikipedia.org../../../r/e/t/Talk%7ERete_algorithm_ee97.html"