Providing flexible access to relevant information in an effective and efficient way remains a challenge. Approaches based on attention metadata present novel opportunities to deal with this challenge. Attention metadata indicate what end users do with information objects. Applications such as authoring tools, search tools, repositories, email applications, blogs and wikis can generate attention metadata. Collecting, managing, merging, analyzing and exchanging distributed attention metadata enables advanced tools to exploit data about user experiences in heterogeneous contexts for more flexible user information interaction. For instance, users of one application can be recommended relevant material, based on their attention given to information that they have interacted with in other applications. More concretely, this approach could allow a user recommendation service to take into account whether the end user is processing her email or authoring a scientific paper while she is listening to the music.