Intelligent Systems - People Search Directory
Download
Report
Transcript Intelligent Systems - People Search Directory
Intelligent Systems
MGMT 661 - Summer 2012
Night #7, Part 1
Outline for Tonight
1.
Intelligent Systems
2.
Decision Support Systems
3.
Enterprise Knowledge Management Systems
Virtual Reality
Neural Networks
Expert Systems
DSS and DSS models
GIS
ESS (e.g. Balanced Scorecard)
Final Homework
tutorial on Pivot Charts
Information Processing
data = recording of events
information = organized data
knowledge = information + (patterns or
context)
wisdom = knowledge + experience
Types of Information
Structured Information
explicitly laid out as tables, graphs, reports
Unstructured Information
emails
presentations
memos
messages
Enterprise-wide Knowledge
Management Systems
contents:
structured and unstructured information
FAQs, work blogs, white papers, special reports
directory of in-house experts
wiki pages
features:
search engines
collaboration tools (e.g. wikis and bookmarking)
automatic knowledge collection
Virtual Reality Systems
Current primary uses:
engineering
design
medical imaging
Future Uses:
immersive
data mining
Virtual Reality Today
Google Goggles
cell phone app
take a picture of a landmark and it tells you where
you are
Google Glasses
hands free smart phone display
responds to voice commands
Neural Networks
used to model complex relationships between
inputs and outputs or to find patterns in data
good at classification problems
learns by example
Expert Systems
Elements of problem solving
facts about the problem
theories about the problem
strategies for solving these types of problems
rules for what to do
Components of an Expert System
Knowledge
facts (known and inferred)
heuristics (rule of thumb)
Inference
Base
Engine
method for reasoning about the facts and heuristics to form a
conclusion
Expert Systems
Example Heuristics
if (temp is cold)
then (need to wear a coat)
if (sky is raining)
then (need an umbrella)
if (need to wear a coat)
and (going outside)
then (put on coat)
and (retract (need to wear a coat))
if (checked temp)
and (checked sky)
and (going outside)
then (go outside)
Expert Systems
well suited to mimicking trouble shooters
interpretation, diagnosis, design, repair, control
able to explain how and why a decision was reached
reasons to develop an ES:
capturing scarce expertise
consistent decisions
faster response
Problems:
expertise
is hard to extract from humans
Group Exercise
Write four expert system rules to
determine which courses an MBA student
should take next semester.
if (fact)
then (new fact)