Light-weight Ontology Editor

At the moment some ontology editors are available to model ontologies in research and industry, e.g., Protegé, TopBraid Composer, and OntoStudio. However, these tools are too complex and offer too much functionality that even experts in semantic technologies can hardly operate them properly. In contrast, our work focuses on developing an easy-to-use and light-weight ontology editor which can be used by non-experts w.r.t. semantic technologies.

General Use Case

The development of the light-weight ontology editor is the following scenario occurring in a company: A company wants to improve some functionality of an application by using semantic technologies. Typically, such a scenario includes the development of an ontology modeling the application domain. Since the development of ontologies is a complex process – comparable to processes in software engineering – a first version of the ontology has to be constructed by ontology engineering experts in collaboration with domain experts. Domain ontologies are generally not static but have to be adapted to changes in the application domain, e.g., ontologies in the chemistry domain have to be modified periodically due to new research results and products. In most cases, the changes at a time will be rather small, e.g., rename a concept or add a new concept, and, thus, the costs of a hiring an ontology engineer will be out of all proportion. As a consequence, an employee of the company should be responsible for keeping the ontology up-to-date. However, in the absence of appropriate tools for non-experts it is currently a challenge for the employee to handle this task.

Balanced Editor Functionality

Besides the functionality of creating and modifying concepts and relationships between them, current ontology editors offer a good deal more than this. For example – to mention only a few – it is possible to create rules, run an inference engine on the ontology, or give meta information on properties (e.g., symmetric, functional). Developing the ontology editor we decided to focus primarily on basic operations such as searching, adding, removing, and editing concepts and relationships. Additionally, we planned to provide tagging and discussing mechanisms to users. Please, see our technical report for a short explanation of these operations.

Key Components

At the lowest level of the architecture an RDF store is responsible for storing and querying the data produced by the ontology editor. Accessing functionality of the component repository and project manager a person with some knowledge in semantic technologies, e.g., an ontology engineer, can setup the ontology editor for the use by domain experts. For example, he has to create and configure a repository containing the projects belonging to an ontology and import the initial ontologies and design patterns. The component design pattern manager supports the administrator of the ontology editor to create, delete, or modify design patterns. Moreover, it generates input forms for inserting data by ordinary users based on a design pattern. The discussion manager handles user discussions about concepts and relationships of the ontologies, keeps track of arguments, and supports the process of decision-making. The discussion process will be structured according to a well-known methodology, e.g., DILIGENT supports controlled discussions besides others. Finally, the component ontology support helps the user to edit ontologies appropriately, e.g., avoiding inconsistencies and redundancy or placing a new concept in the ontology. For example, inconsistencies may occur if a user deletes some concept of a project which is needed by other ones and redundancy if he creates a concept that already exists in the project with a similar label (e.g., "alcohol" vs. its German translation "Alkohol").

Team

Ralf Heese