Scientifically grounded methodology is indispensable to developing complex and modern IT systems, and methodologies are in constant change.

The latest trends develop object-oriented programming into component-based, model-based and aspect-oriented development. To create the most generalisable and recyclable solutions possible, meta models are defined and design templates are created. The practical implementation of methodologies is helped by development frameworks. These frameworks can be continuously upgraded with the modules of new fields of application with the help of meta-models. The developments aim at attaining seamless development technologies, which would eliminate or minimise the current technological and conceptual gap between requirements and models, and source codes and models.

Our most important goal is working out a development methodology based on multi-layer and multi-dimensional meta-models – M(LD)M – and a development framework based on this methodology. The framework is of modular structure and can be upgraded with domain-specific modules based on the unified meta-models. This makes application development for the specific domain fast and efficient.

The insertion of legacy codes is still also to be solved. This expression refers to the problem of integrating older program codes into systems developed using model-based methodology and tools. Legacy codes are often valuable, tried and tested, but have been developed using other methodology and accordingly pose a problem of compatibility.

These objectives designate research tasks such as: developing the methods of domain specific modelling, and making models of the domains connected with the project (document management, traffic logistics, town management, etc.); development of the algorithms and tools of information management; development of the algorithms of data mining for systematic analysis of the data gathered during operation and development to work out the quality features and assessment criteria.