: Nowadays, personalized education is a very hot topic in technology enhanced learning (TEL) research. To support students during their learning process, the first step consists in capturing the context in which they evolve. Users typically operate in a heterogeneous environment when learning, including learning tools such as Learning Management Systems and non-learning tools and services such as e-mails, instant messaging, or web pages. Thus, user attention in a given context defines the Contextualized Attention Metadata (CAM). Various initiatives and projects allow capturing CAMs in a knowledge workers’ environment not only in the TEL area, but also in other domains like Knowledge Work Support, Personal Information Management and Information Retrieval. After reviewing main existing approaches according to some specific criteria that are of main interest for capturing and sharing user contexts, we present in this paper a framework able to gather CAMs produced by any tool or computer...