Efforts towards Combining Graphics, Uncertainty, and Semantics: A Survey

摘要

Modeling the real world is a basic but vital task in computer science. Graphics, uncertainty, and semantics are three key aspects of understanding structure and "why" when handling complex relationships with imperfect or unknown information. The combination of the three aspects can provide a systematic and effective method for modeling the real world. This paper presents a survey of the efforts towards combining these aspects. One branch of the efforts is to combine graphics and uncertainty as probabilistic graphical models (PGMs), and then associate PGMs with semantics. The other branch is to combine graphics and semantics as graph-based knowledge representations, and then add the probability to handle uncertainty. We introduce the models and methods involved in these efforts and discuss the expressiveness, pros and cons of them. Finally, we suggest future work in this domain.