RFBR Grant 08-07-00157
Research and development of methods and tools for semantic identification of specifications of heterogeneous information resources relevant to a scientific problem and their integration in the specifications of the problem at the scientific information systems
Project starting date
Leonid Kalinichenko, Institute of Informatics Problems of RAS, Moscow, Russia
- Institute of Informatics Problems RAS
The project is aimed at development of methods and tools for scientific problem solving. Specifications of the problems are to be defined independently of information resources used for solving of the problems. Heterogeneous information resources are integrated into specifications of a problem to be solved. Resources may be added or removed supporting the changing in time of the needs of scientific research. However this addition or removal does not affect efficiency of the information system.
The main result of the project achieved in 2008 was the architecture of the integration tools intended for semantic identification and integration of specifications of information resources into specifications of a scientific problem. In 2009 methods and tools for integration of specifications of information resources in the specifications of the problem were developed in accordance with the proposed architecture.
The results achieved while developing integration tools include the following. Tools for semantic identification of resources relevant to a problem using metadata search as well as tools for identification of relevant elements of a problem and resources using ontological annotations were developed. The information models Unifier was embedded into the system for semantically motivated construction of transformations of heterogeneous information resourcesí models into the canonical information model. Methods of integration of resource schemas in a problem schema based on the concept of refinement were developed. The integration tools were extended with the instruments for interactive proof of specification refinement. Methods and tools for construction of views over a problemís classes based on GLAV approach were developed. A peculiarity of the approach developed consists in applying of structural transformations expressed as global views (GAV) to get rid of mismatches of problem and resource type specifications and after that in construction of resource specifications as local view (LAV) expressed as a composition of problem classes.
The objective planned for the final 2010 was an application of the developed methods and tools for problem solving. In accordance with this objective a technique for application of methods and tools for semantic identification and integration of specifications of information resources into specifications of a scientific problem was developed. The technique was applied for solving a problem of secondary standards search for a precise calibration in the field of afterglow of gamma-ray bursts defined by experts from Institute of Space Research of RAS. Problem solving was implemented in the subject mediator support system constituting a part of the Russian Virtual Observatory infrastructure.