Research in knowledge engineering is conducted at the Research Laboratory of System Modelling and Software Technologies in the following directions:
- domain-specific modeling languages, model transformations, modeling tool development,
- semantic technologies and tools (e.g. OWLGrEd, ViziQuer), ontology-based data modeling,
- ad-hoc query languages for medical and other domain experts, using both controlled natural language and visual graph diagram primitives,
- advanced graph diagram visualization techniques.
Research in the area of methods and tools for complex system design was started in 1980-ties, with the graphical system modeling tool GRADE (created in cooperation with the German company Infologistik GmbH), still hardly matched for very large graph-like diagram handling.
A generic platform TDA based on models and model transformations has been developed for visual modeling tool building. The tools created on the basis of TDA include several industry-level process modeling tools, as well as visual semantic technology tools OWLGrEd and ViziQuer.
The available tool set for semantic technologies and relational database semantic re-engineering include OWLGrEd (OWL ontology visualization and visual authoring) and ViziQuer (visual ad-hoc queries over RDF data), as well as OBIS (automated ontology-based information system) and RDB2OWL mapping tool.
The current research on modeling languages and modeling tool development focuses on:
- Advanced modeling methods, based on meta-model specialization concepts,
- Domain-specific modeling languages for machine learning, large computing infrastructure descriptions and other applications,
- Development of model-based web platforms for visual modeling,
- Advancement of TDA concepts into creation of webAppOS prototype.
The semantic technology and tool development has its current focus on:
- Visual and interactive notation and ViziQuer tool for ontology-based data analysis queries,
- Advanced ontology visualization techniques,
- Practical data modeling environment creation over RDF and relational data in application areas involving health care, forensics, cultural heritage and public policy making.
The research of ad-hoc query languages for domain experts include:
- Controlled natural language based query approach with practical emphasis on medical data, and
- Visual and interactive ViziQuer notation adaptation for domain expert end users.
Leading researchers: Jānis Bārzdiņš, Kārlis Čerāns, Audris Kalniņš, Agris Šostaks, Elīna Kalniņa, Sergejs Kozlovičs, Edgars Rencis, Lelde Lāce, Artūrs Sproģis, Kārlis Freivalds, Paulis Ķikusts.