On Large-Scale Qualitative Spatio-Temporal Constraint Redundancy Removal

Qualitative spatio-temporal representation and reasoning is an important research direction in the field of artificial intelligence, and qualitative constraint networks are a key tool for representing and reasoning with spatio-temporal knowledge. Previous studies have pointed out that qualitative constraint networks often contain redundant constraints. This not only results in a waste of storage and transmission resources, but can also be a bottleneck in various algorithms and applications that are based on qualitative constraint networks. Existing research proposed several algorithms to deal with redundant constraints, but their time complexities are relatively high, which is very unfavorable when dealing with large-scale data that are common nowadays. This paper proposes a new model based on distributed computation to deal with redundant constraints efficiently. Specifically, this paper theoretically proves the feasibility of the distributed model for this problem, proposes different optimization algorithms for distributed task allocation, and evaluates the performance of the distributed model through a variety of experiments on real-world data sets. The results show that the distributed model can significantly improve the processing efficiency of simplifying representations of qualitative spatio-temporal information on large-scale data sets, and that the distributed solution to the problem of simplifying representations can also improve data privacy, which can contribute to big data processing and data security protection in qualitative spatiotemporal reasoning research.