A matching algorithm for electronic data interchange

  • Authors:
  • Rami Rifaieh;Uddam Chukmol;Nabila Benharkat

  • Affiliations:
  • San Diego Supercomputer Center, University of California San Diego, La Jolla, CA;Computer Science Department, Combodia Technological Institute, Phnom Penh, Cambodia;LIRIS, National Institute of Applied Science of Lyon, Villeurbanne, France

  • Venue:
  • TES'05 Proceedings of the 6th international conference on Technologies for E-Services
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

One of the problems in the actual electronic commerce is laid on the data heterogeneity (i.e. format and vocabulary). This representation incompatibility, particularly in the EDI (Electronic Data Interchange), is managed manually with help from a human expert consulting the usage guideline of each message to translate. This manual work is tedious, error-prone and expensive. The goal of this work is to partially automate the semantic correspondence discovery between the EDI messages of various standards by using XML Schema as the pivot format. This semi-automatic schema matching algorithm take two schemata of EDI messages as the input, compute the basic similarity between each pair of elements by comparing their textual description and data type. Then, it computes the structural similarity value basing on the structural neighbors of each element (ancestor, sibling, immediate children and leaf elements) with an aggregation function. The basic similarity and structural similarity values are used in the pair wise element similarity computing which is the final similarity value between two elements. The paper shows as well some implementation issues and a scenario of test for EX-SMAL with messages coming from EDIFACT and SWIFT standards.