[December 22, 2001] At its Thirteenth Plenary Workshop on October 17, 2001, the CEN/ISSS Electronic Commerce Workshop approved a new project proposal for The Multilingual Upper Level Electronic Commerce Ontology (MULECO). The MULECO project "will research the most efficient means of developing a multilingual upper-level ontology for describing and identifying the relationships between electronic commerce applications and the ontologies used to describe them. In particular it will investigate how information related to business processes can be integrated with existing techniques for classifying businesses, their products and services... The MULECO project will include proof-of-concept as part of the activity."
The MULECO project was proposed by The SGML Centre (Martin Bryan), with support expressed from the following members of the CEN/ISSS Electronic Commerce Workshop: Man-Sze Li (IC Focus), Andrzej Bialecki (WebGiro), Ambjörn Naeve (KTH/CID), Frédéric Camuzard (Motorola), Gerhard Friedrich (UKLA), Zhan Cui (BTexact).
MULECO description from the 2001-12-17 draft (excerpts)
There are many existing and proposed 'electronic commerce ontologies'. The vast majority have been defined monolingually, or in at most three or four languages, often from the same language group. The problem is that different trading partners tend to use different ontologies, and tend to prefer ontologies developed in their native language or in a 'neutral' language, which is often English. It is, therefore, difficult to identify points of overlap between ontologies, and it is also difficult for people to find relevant terms in ontologies using their native language...The aim of MULECO is to develop a mechanism that will allow existing ontologies to identify their inter-relationships by identifying the relationships between themselves and a set of terms defined in a multilingual ontology that has been designed specifically to allow people to find terms using their native language. We realise that it is not possible, or desirable, to create and maintain a multilingual ontology that covers all terms used in all business applications in all European languages. What is needed is a way of classifying entries at the upper-most levels of existing ontologies in a form that takes account of the sort of terms used by people when they are trying to locate the term(s) they wish to use. To do this we need to extend existing business classification schemes to take account of things like business processes, variant names within different user communities, exclusion properties (e.g., no peanuts), etc. Such extensions need to be based on a well documented model that is based on properly researched linguistic characteristics, such as that provided by the Expert Advisory Group in Language Engineering Standards in The EAGLES Guidelines for Lexical Semantic Standards provided in Chapter 6 of EAGLES LE3-4244: Preliminary Recommendations on Lexical Semantic Encoding -- Final Report. The MULECO project will develop an upper-level ontology, expressed as an extended network of industry descriptors, commercial terms and business roles, that will be recorded in a way that allows each entry to be addressed from other ontologies and applications by means of a Uniform Resource Identifier or an XML Path/Query.
The ontology representation language should be expressed in XML so that individual components of it can be referenced as component parts of either a Unique Resource Indicator (URI), XML Path definition or XML Query.
The underlying structure of the XML should be based on the concepts described in the EAGLES framework, but with alternative forms of element names based on typical business renditions of technical terms (e.g., BroaderTerm in place of Hypernym). The terms to be adopted form EAGLES, and their equivalent business terms are shown in the following table [...]
MULECO is an on-going project, and as yet no formal set of definitions, or accompanying DTD/Schema, has been produced. Areas of ongoing study include those currently being undertaken by European research projects such as MILES, CLAMOUR and OntoWeb, and by international e-commerce initiatives such ebXML/ebTWG, related to:
- Formal languages for describing ontologies
- Formal languages for describing multilingual word sets
- Formal models for the maintaining industrial classification schemes
- Formal languages for modelling business processes
- Techniques for the creation and maintenance of process-based ontologies.
The upper-level ontology will take as its start point existing standardized industry and process classification schemes, such as the International Standard for Industrial Classification (ISIC) used as the basis for the NACE classification of European business. The project will take note of the work being done by the IST CLAMOUR project to formally define such classification schemes. In particular it will extend currently used techniques for data classification, based on hierarchical classification of terms into broader and narrower meanings, by allowing for more complex relationships, in particular those relating to the relationships of wholes and parts which are vital to the mapping of the relationships between business processes. By defining a set of business relevant relationships between terms the project will allow classification hierarchies to become a controlled network of related words that forms an ontology rather than a classification scheme.
The project goal is to "research the most efficient means of developing a multilingual upper-level ontology for describing and identifying the relationships between electronic commerce applications and the ontologies used to describe them. In particular to investigate how information related to business processes can be integrated with existing techniques for classifying businesses, their products and services... The group will maintain active liaison with those groups in W3C and ISO responsible for the definition of semantics for electronic commerce applications. It will monitor and comment on the work of the Core Components group of the ebXML initiative, which is concerned with the expression of EDI semantics within XML applications. In addition there will be a liaison to the Foundation for Intelligent Physical Agents (FIPA) who are in the process of developing standards for the use of ontologies by intelligent software agents..." [from the project proposal, referenced below]
References:
- "MULECO -- Multilingual Upper-Level Electronic Commerce Ontology." [Edited] by Martin Bryan. MULECO draft CWA. December 17, 2001. 20 pages. "This CEN/ISSS Electronic Commerce Workshop project will research the most efficient means of developing a multilingual upper-level ontology for describing and identifying the relationships between electronic commerce applications and the ontologies used to describe them." See similarly from the FTP archive; source. [source]
- CEN/ISSS Electronic Commerce Workshop MULECO Project Document List
- MULECO Project Registration form
- MULECO FTP document archive
- Project Proposal: Multilingual Upper-Level Electronic Commerce Ontology. By Martin Bryan. References: ISSS/WS-EC-MULECO/01/001. ISSSS/WS-EC/01/105. Electronic Commerce Workshop Project Proposal. [source]
- Electronic Commerce Workshop CWM [Common Warehouse Metamodel] Business Nomenclature Package. By Martin Bryan. Reference: ISSS/WS-EC-MULECO/01/002. Informational. "The attached document contains screen shots of the model for recording Business Nomenclatures developed as part of the Common Warehouse Metamodel published by OMG. See the 500+ page specification, which can be obtained from OMG: http://www.omg.org/technology/documents/formal/cwm.htm. [source]
- "The Case for a Multilingual Upper-Level Electronic Commerce." By Martin Bryan (The SGML Centre). May 31, 2001. 15 pages. "The classification of business information exchanged via the Internet is of vital importance for the long-term understanding of business processes. This paper looks at how business information can be classified, and the role that ontologies and lexical semantic standards should have in this process. The discussion will illustrate some of the limitations of the current approaches being adopted for the Semantic Web based on the use of the Resource Description Framework (RDF) and the XML representation of Topic Maps (XTM), and suggest ways in which the EAGLES Guidelines for Lexical Semantic Standards and proposed ontology standards such as the Knowledge Interchange Format (KIF) and the DARPA Agent Markup Language Ontology Inference Layer (DAML+OIL) might help to improve existing techniques ... The EAGLES project is concerned with Natural Language Processing (NLP). As such it has a very wide theme, and needs to cater for the large number of circumstances in which text is used. Many of its features are concerned with word disambiguation in different contexts that are not directly applicable to the more limited applications for which business semantics are required. This paper, therefore, only discusses those features of the EAGLES Guidelines that are directly relevant for the description of business semantics..." [source]
- CEN/ISSS Electronic Commerce Workshop
- CEN/ISSS: Electronic Commerce Workshop New Activities
- CEN/ISSS Electronic Commerce Workshop document list
- Electronic Commerce Workshop Contact Persons
- Contact: Martin Bryan (The SGML Centre)
