One way to improve the education of our students is to promote the creation of massive collaborative projects. These projects, also seen as learning products, would also help to better scale our learning experiences (Massive Open Online Courses-MOOCs) and generate collective value from the hours and cognitive efforts invested doing academic work. But the complexities to assess those projects are challenging. This requires to develop new monitoring and feedback systems for these kind of projects. This article presents an exploratory analysis for applying learning analytics methodologies based on social networks analysis, factorial analysis, k-means clustering and “naïve” Bayes algorithms.