Mechanisms are described for providing visual representations of metadata so as to allow for an easier, more intuitive, and more entertaining way for a user to manage metadata by allowing a user to execute operations relating to the metadata through interactions with the visual representations of that metadata. A user can further define visual objects and use such objects for searching media content items, such as by using a visual representation of metadata as a filter and displaying graphical depictions of associations between the visual representations of metadata and the associated media content item. Moreover, operations executed on the visual representations, such as copying, pasting, and modifying the visual representations, are translated to actions on the metadata itself and can be applied across different devices, as well as different media content items and groups of media content items.