Molly Azami

RosettaNet Ontology

• pips.owl version 1.4 - updated 4/25/04

Overview:

The ontology developed for this project is derived from RosettaNet standards and has been written in Owl-DL language. RosettaNet was chosen because of its widespread use and the fact that it has been implemented by many of the big players in the e-commerce arena, such as Cisco, IBM, Intel, Microsoft, Sun and UPS.

Lynda Yana [1999], the Director of Global Marketing and Communications at RosettaNet refers to RosettaNet as an ontology. There is structure to RosettaNet and they makes use of UML diagrams to explain its architecture. RosettaNet is also agreed upon between partners. So in many senses of the word RosettaNet is an ontology. Its use of XML and UML make it a semi-formal ontology. Our efforts in this project are to create a more formal ontology by not only marking up the already existing structure with OWL, but also linking the various transactions and other parts of the RosettaNet messaging system by formally defining relationships.

RosettaNet standards are built on their Partner Interactions Processes (PIPs). PIPs are categorized into seven Clusters, or groups of core business processes, that represent the backbone of the trading network. Each cluster is broken down into segments, cross-enterprise processes involving more than one type of trading partner. Within each Segment are individual PIPs. [RossetNet.org]

Each PIP is made up of least two messages, one to initiate the transaction (the outgoing message) and one as the response to the initial message (the incoming message). Each message is made up of a numerous elements each defined by its cardinality and description in the RosettaNet specifications.

Some of the elements within an individual message are full documents in their own right, like quotes or purchase orders. These documents, while being made up of numerous elements themselves, carry a unique Global Document Identification Number which is used for tracking and processing with in the transaction.

The ontology for this project (pips.owl) was built by traversing the specifications for each message to determine the classes, cardinality, relationships and properties within the OWL ontology.

Beyond just the incoming and outgoing messages of each PIP RosettaNet provides the structure for a complete transaction conversation which includes business signals. The business signals are of two types; receipt acknowledgements and exceptions.

Figure 1 is a high level representation of how the PIP "Request Quote" would be implemented using our pips.owl ontology.

The first phase of creation of the ontology was a purely manual process and only captured those elements within each message that are required (cardinality of at least 1, 1...n). This first draft can be used as the foundation for building a full RosettaNet ontology for further and more complete semantic web processes.

Related Work:

[Zhao 2003] has developed a "vocabulary capture" method in her research on reusing existing standards for ontologies in Internet commerce. In her project the vocabulary from RosettaNet specifications is captured and used for various parts of a web service/web process ontology. In particular RosettaNet wording is used for messaging, syntax & semantics, and vocabulary. The ontology itself is built based on the services the developer wishes to supply. Our pips.owl ontology is built specifically on RosettaNet using not only its vocabulary, but also the PIP architecture and structure of the exchange between partners.