Transcript AIAI Presentation

<I-N-C-A>
A Shared Model for Mixed-initiative Synthesis Tasks
Austin Tate
Artificial Intelligence Applications Institute
University of Edinburgh
http://i-x.info
Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
1
I-Technology
I-P2
Process
Panels
I-DE
Domain
Editor
Cooperation and
Communication
I-Q
Adaptor
Other
Agents &
Services
CoABS Grid,
KAoS,
AKTBus,
XML Sockets,
Jabber,
[Globus GT3]
I-Plan/O-Plan
Planning
Aid
Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
2
I-X Approach



The I-X approach involves the use of shared models for
task-directed communication between human and
computer agents who are jointly exploring (via some
process) a range of alternative options for the synthesis
of an artifact such as a design or a plan (termed a
product).
I-X system or agent views the product as being
represented by a set of constraints on the space of all
possible products in the application domain.
I-X system or agent is viewed as having two cycles:
– Handle Issues
– Manage Domain Constraints


Agents can use their capabilities to handle issues or
carry out given tasks where “authorised” to do so.
Mixed initiative model for each agent of
“mutually constraining the space of products”.
Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
3
<I-N-C-A> Ontology
Issues
Nodes
E.g. activities in a process or parts in a physical artifact
Constraints
Critical Constraints (shared across multiple components)
Auxiliary Constraints (localised to a single component)
Annotations
E.g. decision rationale and other notes
Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
4
Uses of a Shared Model
Knowledge
Acquisition
User
Communication
< I-N-C-A>
Formal
Analysis
System
Manipulation
Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
5
Terminology to Support
Mixed-Initiative Working

Processes -- <I-N-CA>
Products -- <I-N-CA>
Options for Products
Levels within Product Descriptions
Parts (or Phases) within Product Descriptions

Example use in TRAINS   O-Plan




Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
6
O-Plan Web Server Multi-user
Task & Planning Support
I-X Process Panels
Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
8
Aim is a Workflow and
Messaging “Catch All”

Can take ANY requirement to:
–
–
–
–

Deals with these via:
–
–
–
–
–

Handle an Issue
Perform an Activity
Respect a Constraint
Note an Annotation
Internal capabilities
External capabilities
Manual activity
Reroute or delegate to other panels or agents
Plan and execute a composite of these capabilities
Receives reports and interprets them to:
– Understand current status of issues, activities, constraints & annot.
– Understand current world state, especially status of process products
– Help user control the situation

Copes with partial knowledge
Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
9
I-X Process Panels
Process Panel
Activity Editor
Domain Editor
Map View
PDA View
Messenger
I-Space
3D View
Further Information


http://www.aiai.ed.ac.uk/project/ix/
http://i-x.info
Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
12
Tasks & Initiative

Strategic
–
–
–
–

Tactical
–
–
–
–

Tasks: objective setting, option analysis and decision making
Task Types: analysis, direction
Aim: What to do?
Initiative (in many of our scenarios): Often manual
Tasks: planning, scheduling, option generation
Task Types: synthesis
Aim: How to do?
Initiative: Greatest opportunity for mixed-initiative
Operational
–
–
–
–
Tasks: Enactment, adjust approach in context, select from options
Task types: execution, selection, modification
Aim: Just do it!
Initiative: Some parts may be automated
Artificial Intelligence Applications Institute
Centre for Intelligent Systems and their Applications
13