External layer

Knowledge representation

Reports

Data entry and -manipulation forms

Query forms

Application files

Charts

XML/XSD files

HTML pages

Browsers

Utilities

Conceptual layer

Knowledge management

Physical layer

Knowledge base

RDBMS

HTML pages

Flat text files

Other ontologies' individuals

ODBC/JDBC connections

XML files

ANSI/SPARC model

Standard information model divided in functional layers

Application files

Besides the files mentioned under reports, these comprise OpenOffice Drawings or calculation sheets, Excel, Access, .dbf and other application related files

Browsers

Browse through data (called individuals) and knowledge.

Charts

Output summarizing and presenting data in a visual way but also the knowledge representation around processes and procedures.

Conceptual layer

The layer which defines concepts and relates them to each other for a specific domain.

When this layer is missing IT and business departments work beside each other without direction; with loss of energy, efficiency, operational control, uncontrolled risks.

Insufficient product, process and procedure knowledge to manage a domain is the intermediary stage to disasters.

For more information we refer to the business case pages.

The main users are business analysts and process managers.

Data entry and

-manipulation forms

Domain-user interface to enter new data or manipulate existing data.

Those forms always contains a filtering function.

External layer

Is the layer of the users of the domain's knowledge.

This ranges from the top-management, via the front-desk employee to the mainframe programmer.

HTML pages

Web pages

Knowledge model

Organization of knowledge in a domain

ODBC/JDBC

Open (MS) or Java (Sun) Data Base Connections

The techniques allow accessing data in proprietary databases without having to install the whole management system.

Ontologies' individuals

Instances of classes in an ontology are called individuals.

They can be accessed through their URI and be used in other ontologies. E.g. the finance ontology contains several ISO-standard related individuals (stock markets, currencies, countries) as well as the UN standard for activity codes (ISIC). The latter are often referred to as NACE-codes in the EU.

Ontology

Management system of knowledge.

Ontologies can be organized in a way, particular to the supplier of the information system.

Others can follow standards.

Ontologies allow a flexible adaptation to new/changed realities through their atomized concepts.

RDF (Resource Description Framework) is such a W3C standard.

RDF dresses classes, individuals and properties with URI's (Uniform Resource Identifiers) which refer to physical information.

OWL (Web Ontology Language) is build on top of RDF and versions 1.0, 1.1 and 2.0 are W3C standards.

OWL exists in 3 flavors: lite, DL (description logic) and full.

The light and DL versions comply with restrictions so that reasoners can infer knowledge of the asserted statements.

RDF and OWL can both be accessed through the internet network technology (local or www).

PCPACK

Current version is 5. Version 6 would become available in July, 2009.

The ontology format is proprietary (XML). From version 6 on, import and export to RDF and OWL format would be available.

This tool is ideal when real-time updates to RDF and OWL formats are not necessary because of the very good knowledge representation means.

Those happen through a user-friendly website by means of forms, knowledge representation charts,...

The efficient knowledge representation means have to be leveraged against:

- the lack of individuals information, caused by the file format;

- no support to keep the ontology into OWL-DL limits so that reasoners can be used.

The website is not that technical, servers are necessary. Just a separate folder on the network (or on the web) is sufficient.

The threshold to start using the package is low with very good documentation and tutorial. In many circumstances this solution can be a very good alternative for a not too heavy knowledge management system.

I have tested it (not stress tests) and was really happy.

The manufacturer of the tool is Epistemics.

Physical layer

Layer where the information is found physically.

The organization and management of that layer is the work of IT-people mainly, except for public access data (ontologies, html, xml,...). There the role of IT-departments can be restricted to providing access to and description support for the data.

Protégé

There are 2 versions of Protégé developed by the university of Stanford (Medical department) and the university of Manchester for respectively versions 3.4 and 4.0.

Both versions are necessary if one wishes to take benefit of the many plugins developed for both versions. They also support different versions of OWL. Both support RDF but 3.4 supports many other formats, while 4.0 supports OWL 2.0.

Protégé is a very good application to start developing ontologies because of the good tutorials, support from user groups and plugin developments.

Links for Protégé 3.4 and Protégé 4.0.

Query forms

Domain-user interface to specify the data which will be used in output.

Advanced queries can select the data and specify the manipulation to execute.

RDBMS

Relational Data Base Management Systems like MS Access, Oracle, MySQL,...

Reports

Lists of data available to the domain-user in any format: .doc, .rtf, .html, .pdf, .txt,... Those lists can eventually be filtered, printed and saved depending on the applications linked to the formats.

Tools

Software tools available to fulfill the tasks specified by the models

TopBraid

The Maestro edition is a suite of tools which is tested to the great satisfaction of every-one who came into contact with it.

We may conclude that finally the semantic data (via the web or otherwise) era has begun.

It is a professional package, including application development environment and running on the Eclipse software development platform.

The latter eases proprietary developments around the tool.

The threshold is relatively high: one has to know what he is doing, but the help is correct.

Many ontology-related tools are included: SPARQL, SWRL rules, integration of XML/XSD,...

The manufacturer is TopQuadrant.

For the finance ontology, we choose TopBraid as the development and deployment platform.

To that effect, and for the purpose of offering development support services around the package, we entered into a Solution Partner Agreement on September, 22nd 2009.

Utilities

Representation techniques of output. E.g. hyperbolic tree, jOWL,...

See utilities page

Virtuoso

We have chosen OpenLink's Virtuoso as our preferred implementation partner. Find out the reasons.

XML/XSD files

Message files and message format specification files.