Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/2548
Title: Product schema integration in an agent-based electronic commerce system
Authors: Yan, Guanghao
Keywords: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
Issue Date: 2000
Abstract: The ABECOS (Agent-Based Electronic COmmerce System) project investigates issues in building an infrastructure for a large scale, distributed, agent-based electronic commerce system in which buyer and seller agents liaise with one another to facilitate electronic commerce transactions. In such a system, a buyer agent is interested in purchasing a product from one or more seller agents. However, the heterogeneity of product descriptions is an impediment to successful interactions among agents. Only when agents have a common ontology of product description structures (also called product schemas) are they able to engage correctly in electronic commerce activities. This gives rise to the problem of Product Schema Integration (PSI); to integrate heterogeneous schemas of a certain product type into one globally compatible schema. In this work, we address this problem in the context of the ABECOS system. We divide PSI into two sub-problems set schema integration and tree schema integration, and present a formal definition of each sub-problem and show that they are NP-complete. Then, we propose a set of approximation algorithms for PSI and evaluate their performance experimentally. In addition, we address issues in integrated schema maintenance andx
URI: http://hdl.handle.net/10356/2548
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:SCSE Theses

Files in This Item:
File Description SizeFormat 
SCE-THESES_292.pdf
  Restricted Access
11.27 MBAdobe PDFView/Open

Page view(s) 5

396
checked on Sep 28, 2020

Download(s) 5

5
checked on Sep 28, 2020

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.