Mining at Most Top-K% Spatio-temporal Outlier Based Context: A Summary of Results

Discovering STCOD is an important problem with many applications such as geological disaster monitoring, geophysical exploration, public safety and health etc. However, determining suitable interest measure thresholds is a difficult task. In the paper, we define the problem of mining at most top-K% STCOD patterns without using user-defined thresholds and propose a novel at most top-K% STCOD mining algorithm by using a graph based random walk model. Analytical and experimental results show that the proposed algorithm is correct and complete. Results show the proposed method is computationally more efficient than naive algorithms. The effectiveness of our methods is justified by empirical results on real data sets. It shows that the algorithms are effective and validate.