Recent years have seen a proliferation of technological systems designed to allow for collaboration and communication between groups of people (electronic messaging systems, social networks, shared workspaces, etc,). Within this context, there emerges the challenge of analyzing the work processes supported by the collaborative systems in order to improve the experience of the participants and to adapt the support offered by these systems to the specific needs of each case. Analyzing collaboration and interaction within the processes of group work is a complex task which requires many issues to be resolved, such as what information is to be processed, how variables for analyzing the group work and the resulting products are to be inferred, and how the results of the analysis can be used to improve the users' activity. To facilitate these tasks, the work that we present here focuses on modeling and automating the collaboration and interaction analysis in order to better understand how user groups orchestrate their activity and to evaluate the products that result from the work that they carry out. The research hypothesis of this thesis proposes that it is feasible to design and build a framework which provides software developers with a guide and an effective support for the construction of systems that automate the analysis - whether separate or combined- of the collaborative process and, in the form of artifacts or products, the results of this process. In order to validate this research hypothesis, the main objective of this project is to build just such a framework, made up of a conceptual framework and a technological framework, so as to facilitate the automation of the collaboration and interaction analysis tasks. The conceptual framework consists of a set of ontologies that define and explicitly relate all of the elements involved in the collaboration and interaction analysis. The technological framework includes a computational support that enables developers to follow a model-driven approach to carry out the analysis processes automatically. This computational support is functionally structured into three levels: (i) the meta-information level, where the analysis processes are modeled using visual tools and the work processes to be studied are specified; (ii) the analysis level, where the models specified in the previous level are processed, combined and transformed to create a computational support that allows for the automatic execution of the analysis processes; and (iii) the collaboration level, where interaction with the collaborative system occurs in order to capture the actions and products generated in the course of the collaborative activity, to calculate variables for analyzing the collaborative process and, ultimately, to improve the users' work according to the results of the analysis. In order to evaluate and validate the proposed framework, several theoretical and practical studies with real users have been carried out to analyze collaboration and interaction within different work scenarios, which differ in terms of the type of collaborative system used and the nature of the tasks performed by the participants. Likewise, we have empirically studied how some automated analysis processes lead to an improvement in users' activity. The results of these studies have made it possible to validate the formulated research hypothesis since the proposed framework has enabled the automation of the required analysis tasks and the various participants have provided a positive evaluation of the support provided by the framework.