Ontology-Based Product Identification to Support Electronic Commerce

Project Summary

The objective of this project is to provide network-based assistance with the task of identifying products which satisfy a user's requirements and are available for purchase. The context for this project is provided by the U.S. Federal Government's Defense Logistics Agency (DLA). The primary focus is on building an ontology-based flexible item identification system that will enable buyers to determine the National Stock Numbers (NSNs) of items in DLA's Federal Supply Catalog (FSC) that satisfy their requirements. The system will provide a significantly richer item description language than is now available in DLA systems, including a functional description language, and will provide both a hierarchical browsing and a query-based search user interface. The ontology-based flexible item identification system will contain an item description ontology that elaborates the terms in the FSC and Federal Item Identification Guide and includes class-subclass hierarchies, part-subpart hierarchies and function-subfunction hierarchies. The item description ontology will be in a form that will enable it to be used by multiple modules for multiple purposes.

The result of this effort will have numerous payoffs. The ontology-based flexible item identification system will ease the task of finding items of supply by:

Supporting hierarchical browsing along multiple dimensions (e.g., class-subclass, part-subpart, function-subfunction);
Providing a richer query vocabulary, including functional descriptions;
Enabling browsing to closely related items retrieved by query (e.g., superclass, subclass, siblings, part-of).

The item description ontology and associated ontology development tools will:

Ease the task of describing new items by enabling top-down semi-automatic classification of new items, and
Ease the task of adding new NSNs by enabling identification of gaps and overlaps in existing descriptions.

This project is being performed in collaboration with the University of Southern California's Information Sciences Institute and with Stanford University's Logic Group.