Methods, systems, apparatuses, and computer program products are provided for selecting content on a displayed page. A selection mode may be initiated with regard to a displayed document that includes content. The initiation of the selection mode may be detected. Multiple entities in the displayed document are determined, with each entity including a portion of the content of the document. The displayed document is annotated to indicate the determined entities. Content selected by a user in the displayed document may be detected. At least one entity is determined to be associated with the selected content. The determined associated entity/entities are indicated in the displayed document as active entities. The user is enabled to refine the set of active entities. An action is enabled to be performed on the active entities.