====================================
REPORT FOR OCTOBER 2007 - MARCH 2008
====================================

TECHNICAL ACCOMPLISHMENTS

() Extended the mechanism for identifying clusters of similar rooms,
which improves the learning of default room properties.

() Developed a mechanism for the dynamic learning of typical resources
and schedule constraints, which enables the system to improve its
"default assumptions" in the process of elicitation.

() Developed an advanced mechanism for the learning of common-sense
assumptions about resources and constraints, represented by "if-then"
inference rules.

() Developed a mechanism for the learning of the correspondence
between conference needs and available goods and services, which
improves the effectiveness of the automated vendor-order selection.

() Developed an initial mechanism for the learning and automated
improvement of elicitation strategies.


IMPROVEMENTS TO PROTOTYPE

() Integrated four of the five new learning mechanisms with the rest
of the system; the only exception is the learning of common-sense
assumptions, which has not yet been integrated. Also, integrated
Vendor Optimizer with ST Module.

() Implemented continuous backup of the state of ST Module, which
allows restoring of the state after crashes and graceful shutdowns.

() Completed the work on the extended GUI for constructing STP world
models.


PROJECT PLANS

() Complete the work on the new learning components of ST Module, and
evaluate their impact on the scheduling and elicitation.

() Complete the integration and debugging of the Year-Four version of
ST Module.

() Prepare a journal paper on the new learning results.


OBJECTIVES FOR NEXT PERIOD

Objective name: Elicitation learning
Objective type: Optional learning
Description: Extend the mechanism for the learning of elicitation strategies
Impact: Automated improvement of elicitation strategies, which will
  reduce the number of elicitation questions

Objective name: Common-sense learning
Objective type: Optional learning and integration
Description: Complete the work on the learning of common-sense assumptions
  and integrate it with the rest of the system
Impact: Improvement of the schedule quality and reduction in the
  number of elicitation questions

Objective name: Selection of room requests
Objective type: Integration
Description: Complete the work on the advanced search mechanism for the
  integration of ST Module with CM Radar, based on the analysis of
  possible schedule improvements after specific room requests, and
  integrate it with the rest of the system.
Impact: Reduction in the number of room requests during the scheduling

Objective name: Placement of vendor orders
Objective type: Integration
Description: Develop a mechanism for the automated placement of vendor orders
Impact: Automated selection and placement of vendor orders with minimal
  user participation, which will reduce manual effort required for order
  placement and improve the schedule quality

Objective name: Testing and debugging
Objective type: Testing and debugging
Description: Thoroughly test and debug the new integrated version of ST Module
Impact: Improvement of the system stability


PUBLICATIONS

We have submitted three conference papers describing the new learning
components; however, these papers are still under review.