Systems and methods of media stream generation based on categories of user expression are disclosed. In an embodiment, a method includes receiving input selecting at least one category from a plurality of categories of expression. Each of the plurality of categories is descriptive of a type of expression made by users as media content is presented. The method includes selecting a group of segments of the media content. Each of the segments of media content in the group is associated with the selected category. The method further includes generating a stream of media content that includes each of the segments of media content.