|
|
 |
|
|
|
|
Scientifical & Technical Short Description
|
|
| | |
|
Collaboration of distributed knowledge communities is a rapidly increasing application field, e.g. international enterprises, scientific research teams, e-learning communities. For efficient collaboration the common understanding of information is a decisive factor. A systematic approach to gain this common understanding is the dynamic creation of ontologies, leading to a more efficient use of shared information resources. At present, the creation of high-quality ontologies is a very time consuming and expensive task. Therefore, such ontologies are available only for few thematic fields, where there is hope for significant economies of scale, e.g. in the health sector, in tourism and in the insurance sector. What is still missing is a methodology supported by tools, which would enable domain experts (who are not ontology building experts) to create ontologies on the fly, yet based on sound principles, created in short time. "Dynamic" means that these ontologies can then be extended and refined over time, possibly by other non-IT experts, can evolve to become more axiomatised, and can be personalised and localised by individuals or groups without losing touch with the community's preferred interpretation. At present, none of the above requirements are sufficiently supported by methodologies and tools. This leads to many poorly analysed "ontologies" being published on the Web. This is likely to become a trust problem and an economic hurdle for the "semantic" Web.
In order to overcome the current bottlenecks, the proposed project DynamOnt offers:
1. to develop a methodological framework for dynamic generation and maintenance of ontologies, including the gradual axiomisation process leading from existing lexico-terminological resources to formalised models
2. to support and enhance the methodology with findings from the research field of formal ontology,
by integrating foundational ontologies such as SUMO and DOLCE.
3. to develop process models that enable user communities to personalise ontologies within common conceptual workspaces
4. to align the work done in ontologies and object oriented modelling with related work in terminology and lexical semantics (e.g. the WordNet project) in order to adequately reflect semantic richness and complexity of knowledge resources
5. to evaluate and validate the methodological framework, in three environments, where communities of practice need to develop their common understanding based on resources ranging from simple project glossaries to formal models of some subject matter.
For the actual development and integration of domain specific ontologies a methodology will be created that is suitable for use by non-ontologists. Experts of the thematic fields (e.g. medical staff, engineers, teachers/students) will be enabled to create ontologies with much less assistance of ontologists than is currently the case. By this a sustained use of content resources will be enabled. The methodology to be developed is a precondition for the creation of sound ontologies in many different thematic fields and communities and this will multiply the benefits of the proposed project.
By bringing together two universities and two non-university research institutions in a single project, DynamOnt offers to develop an Austrian centre of excellence in a research field where there is only a handful of globally renowned players and where there is dramatic industrial growth potential if the European vision of becoming a world-leading knowledge based economy turns into reality.
|
|
|
|
