The key role of models in any Model-Driven Engineering (MDE) process, which principles are to enhance the role of models and to increase the level of automation all along the development process provides a new landscape for dealing with traceability.

In the context of MDE traces are merely links between the elements of the different models handled along the software development cycle. Traces can be consequently stored in models that can be processed by means of model-based techniques. To take advantage of this scenario, this dissertation introduces iTrace, a framework for the management and analysis of traceability information in MDE projects.

To address the shortcomings detected in the state-of-the-art regarding the interchange of traceability information, iTrace brings trace generation mechanisms (both Model-to-Model (M2M) and Model-to-Text (M2T)
transformations) to transformation engines that do not support the generation of traces, or which, do not support so explicitly. In addition, iTrace normalizes existing ad-hoc trace models by bringing them into a common metamodel.

Besides iTrace provides full support to all CRUD operations and a solution for the visualization of trace links by offering two visualization mechanisms: a low-level editor for M2M and M2T traces and a set of dashboards that extracts high-level information from the raw traceability data. This idea is actually brought from the business world, where operational data is used to produce relevant information.

This dissertation presents the methodological proposal bundled in the framework along with the toolkit that supports it. Besides, the validation of the functionality offered by the proposal, as well as the assessment of different features, such as performance or scalability, are also presented.