In interactive multimedia presentations, users should have the flexibility to decide on various scenarios they want to see. This means that two-way communications should be captured by the conceptual model. An abstract semantic model, the augmented transition network (ATN), is proposed for modeling user interactions in multimedia presentations. In ATNs, each state node allows multiple outgoing arcs to model potential user interactions. At the decision point, the multimedia presentation system can use this information to display selection buttons so users can make their choices. The superiority of modeling user interactions with ATN instead of the Timeline model or the Object Composition Petri Net model is discussed in this paper. Our results show that the ATN is effective for modeling user interactions in a multimedia presentation environment.