Cyc

Swimmer963
Swimmer963 (+511/-101)
Alex_Altair (+15)
Alex_Altair (+278/-5) /* Criticism */
Alex_Altair (+2810) Created page with "{{wikilink}} [[File:Cycorp.gif|130px|thumb|right|The official logo of [http://www.cyc.com/ Cycorp], the developer of Cyc. ]] '''Cyc''' (from 'encyclopedia') is a large AI projec..."
The official logo of Cycorp, the developer of Cyc.

 (#$(#$isa #$DouglasLenat #$AIResearcher)

(#$implies
 (#$  (#$and   
 (#$    (#$isa ?OBJ ?SUBSET)?OBJ ?SUBSET)
 (#$    (#$genls ?SUBSET ?SUPERSET)?SUBSET ?SUPERSET))
 (#$  (#$isa ?OBJ ?SUPERSET)?OBJ ?SUPERSET))

(sandcatsandcat ∈ FelineFeline) ∧ (FelineFeline ⊂ MammalsMammals) ⇒ sandcatsandcat ∈ MammalsMammals

  • Lenat, Douglas; Guha, R. V. (1990). Building Large Knowledge-Based Systems: Representation and Inference in the Cyc Project. Addison-Wesley. ISBN 0-201-51752-3.
  • Lenat, Douglas; Miller, George; Yokoi, Toshio (November 1995). "CYC, WordNet, and EDR: Critiques and Responses". Communications of the ACM38 (11): 45-48.
  • Jared Friedman (October 2003). "The Sole Contender for AI". Harvard Science Review.

Many AI researchers criticize the Cyc project for its lack of theoretical foundation. WhileFor example, while the inference engine can perform deductive logic on the fact in the database, it cannot perform induction; that is, it cannot take new data and create new concepts or relations. Furthermore it remains 'crystalline' in the sense that statements are not evaluated probabilistically. Lenat has responded to some critiques by pointing out that Cyc is not trying to build a full AGI, and that it may be easily imported into other AI projects.