Our research group is interested in various topics
- theoretical aspects of software metrics
- practical metric supporting software testing and validation
applications of Information Systems, especially in the
field of Data Warehouse.
We studied the main proposals for software metrics and compared them in large experimental environments.
Various models of new predictive metrics have been proposed, based on structural aspects of software and their relations to statistics of notified faults on software modules. They produce a risk-hierarchy of structural parameters and, using a statistical models, build a frame to identify fault-prone software models. We studied the use of this predictive metrics in the testing and validation phase, due to their capability to identify dangerous constructs of programs on fault-prone modules. The methodology introduced will be able to concentrate the efforts of testing and validation on the more dangerous segments of code or modules.
We are also studying methodologies of Reuse Engineering and their impact in software factories and industrial projects. At last we are analyzing the use of Information Retrieval tools for Maintenance and Reuse of Software Packages.
Regarding Information systems, we studied the effects of data-quality on Data warehouse applications, especially the reliability of related decision-making process.