Symbolic Decision Tree for Interval Data—An Approach Towards Predictive Streaming and Rendering of 3D Models

3D content streaming and rendering system has attracted a significant attention from both academia and industry. However, these systems struggle to provide comparable quality to that of locally stored and rendered 3D data. Since the rendered 3D content on to the client machine is controlled by the users, their interactions have a strong impact on the performance of 3D content streaming and rendering system. Thus, considering user behaviours in these systems could bring significant performance improvements. To achieve this, we propose a symbolic decision tree that captures all attributes that are part of user interactions. The symbolic decision trees are built by pre-processing the attribute values gathered when the user interacts with the 3D dynamic object. We validate our constructed symbolic tree through another set of interactions over the 3D dynamic object by the same user. The validation shows that our symbolic decision tree model can learn the user interactions and is able to predict several interactions with very limited set of summarized symbolic interval data and thus could help in optimizing the 3D content streaming and rendering system to achieve better performance.