1 / 4

Scone Knowledge Base

School of Computer Science. Scone Knowledge Base. Scott E. Fahlman April 14, 2005. Goals for Scone. Build a practical “knowledge base” (KB) engine and toolkit that can be used in any number of applications. Semantic network model: nodes are entities, links are statements.

ethel
Download Presentation

Scone Knowledge Base

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. School of Computer Science Scone Knowledge Base Scott E. Fahlman April 14, 2005

  2. Goals for Scone • Build a practical “knowledge base” (KB) engine and toolkit that can be used in any number of applications. • Semantic network model: nodes are entities, links are statements. • Emphasis on scalability and ease of use. • A few million entities and relations on a high-end workstation. • Parallelizable if you want more. • Implemented in Common Lisp, runs on 32-bit or 64-bit workstations. • Will release as open-source, with a tutorial book. • Alicia Tribble (Ph.D. student) working on NL input of knowledge.

  3. Current Cabilities • Multiple-inheritance hierarchy with exceptions. • Efficient detection of type violations. • Mapping of concepts to words is many-to-many. • User-defined roles and relations. • Can attach meta-knowledge to any node or link. • Multiple contexts: hierarchical & lightweight. • Can represent numerical weights and probabilities. • Programmer-level document. • Current “core” KB is minimal. • Machinery for episodic memory exists, but is minimal.

  4. Multiple Contexts • Each link (statement) in SCONE is tied to the context in which it is true. • Contexts form an inclusion hierarchy. Each starts as a clone, then adds or cancels some information. • To activate a context, mark it and all included contexts, then mark all active statements. Only these link are active. • So multiple contexts are efficient, lightweight, and heavily used for reasoning about alternative states of the universe, real or hypothetical. (“What if…”) • The multiple-context mechanism is used to implement actions and events.

More Related