EDI involves exchange of transaction sets (a sequence of standardized messages) taken from a predetermined set of message types. Each message is composed, according to a standardized syntax, of a sequence of standardized data elements.

Implementation of EDI requires the use of a family of interrelated standards for: (a) the syntax used to compose the messages and separate the various parts of a message, (b) types and definitions of application data elements, most of variable length, (c) the message types defined by the identification and sequence of data elements forming each message, and (d) the definitions and sequence of control data elements in message headers and trailers [FIPS, 2000].

EDI requires a number of software applications such as mapping and translation software. Mapping software is used to establish a relationship between the business application data elements and the EDI standards. Translation software is used to convert the data to an EDI standard from any other format. The mapping software feeds into the translation software so the translation software knows how to convert the data into an EDI transaction set.

Transfer of business transaction sets via Electronic Data Interchange (EDI) usually occurs in the following manner [GEIS, 2000; Frito-Lay, 2000].

1. Preparation of Electronic Documents: The sending organization (sender) organizes the data to be sent using its own business application.
2. Outbound Translation: This group of data is translated into an EDI standard format (transaction set) by the sender.
3. Communication: The transaction set is transferred to the receiving organization (receiver). This transfer of data occurs either through a value-added network (VAN) or through a direct dial-up connection to the receiving organization.
4. Inbound Translation: The transaction set is translated by the receiver into data that is usable by the receiver's business applications.
5. Processing Electronic Documents: The receiver's business applications use the data to conduct business.