Multi-System Report Integration Using Blackboards

From Tetherless World Wiki

Jump to: navigation, search

Citation: John R. Delaney. (1986) Multi-System Report Integration Using Blackboards. In KSL-86-20, 1986.

Publication techreport ( Edit )
type Technical Report
bibtype techreport
Bibtex basics
author John R. Delaney
title Multi-System Report Integration Using Blackboards
number KSL-86-20
institution Knowledge Systems, AI Laboratory
year 1986
Bibtex more
Access Paper
abstract Blackboards are an AI problem solving methodology. A blackboard system consists of a structured data base (the blackboard) holding input and derived inferences and a collection of procedures for deriving inferences (knowledge sources). Each knowledge source is specialized to operate on some portion of the blackboard. The knowledge sources are invoked opportunistically as the information on the blackboard increases.The best known applications of the blackboard methodology have been in speech understanding and passive sonar data interpretation. The inputs in these cases were a single form of raw sensor data. But the methodology is also well suited to integrating multiple streams of fully reduced and qualitatively different data such as active radar track reports, passive electronic intelligence reports, and human intelligence reports about enemy intentions.This paper sketches the nature of the blackboard problem solving methodology with an emphasis on those features suiting it to such applications. The sketch is illustrated with examples from a relatively simple multi-system report integration problem. Relevant applications currently under development at Stanford's Knowledge Systems Laboratory are also described.

KSL Technical Report ID: KSL-86-20
Facts about Multi-System Report Integration Using BlackboardsRDF feed
Abstract Blackboards are an AI problem solving meth … Blackboards are an AI problem solving methodology. A blackboard system consists of a structured data base (the blackboard) holding input and derived inferences and a collection of procedures for deriving inferences (knowledge sources). Each knowledge source is specialized to operate on some portion of the blackboard. The knowledge sources are invoked opportunistically as the information on the blackboard increases.The best known applications of the blackboard methodology have been in speech understanding and passive sonar data interpretation. The inputs in these cases were a single form of raw sensor data. But the methodology is also well suited to integrating multiple streams of fully reduced and qualitatively different data such as active radar track reports, passive electronic intelligence reports, and human intelligence reports about enemy intentions.This paper sketches the nature of the blackboard problem solving methodology with an emphasis on those features suiting it to such applications. The sketch is illustrated with examples from a relatively simple multi-system report integration problem. Relevant applications currently under development at Stanford's Knowledge Systems Laboratory are also described. dge Systems Laboratory are also described.
Author John R. Delaney  +
Bibtype techreport  +
Has author John R. Delaney  +
Has identifier KSL-86-20  +
Has publishing details 1986  +
Has title Multi-System Report Integration Using Blackboards  +
Has where published KSL-86-20  +
Has year 1986  +
Institution Knowledge Systems, AI Laboratory  +
Ksl tr id KSL-86-20  +
Number KSL-86-20  +
Process note YES  +
Title Multi-System Report Integration Using Blackboards  +
Year 1986  +
Retrieved from "http://tw.rpi.edu/wiki/Multi-System_Report_Integration_Using_Blackboards"
Personal tools