Talk:Structure mapping engine
 | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
| |||||||||||
.jpg){kind=small}
This description of SME is somewhat out of date
[edit]We stopped using dempster-shafer in 1989, because simulation results indicated it didn't do as well as trickle-down
Forbus, K. and Gentner, D. (1989). Structural evaluation of analogies: What counts? Proceedings of the Cognitive Science Society.
The current algorithm uses a greedy algorithm to produce approximate best answers in polynomial time. This and other changes are summarized in Forbus, K. D., Ferguson, R. W., Lovett, A., and Gentner, D. (2016). Extending SME to handle large-scale cognitive modeling. Cognitive Science, DOI: 10.1111/cogs.12377, pp 1-50., which can be found online.
I'm willing to do these edits, but since I'm one of the authors, my impression is that such self-edits are frowned upon, or at least need extra vetting. If someone is up for please editing it, great, but if nobody steps up in a month or so, I'll update it. Conser~enwiki (talk) 18:21, 3 February 2024 (UTC)