Bounding Box Representation of Co-location Instances for L∞ Induced Distance Measure
[ 1 ] Instytut Informatyki, Wydział Informatyki i Telekomunikacji, Politechnika Poznańska | [ P ] employee
2021
chapter in monograph / paper
english
- co-location
- bounding box
- data mining
EN In this paper, we investigate the efficiency of Co-location Pattern Mining (CPM). In popular methods for CPM, the most time-consuming step consists of identifying of pattern instances, which are required to calculate the potential interestingness of the pattern. We tackle this problem and provide an instance identification method that has lower complexity than the state-of-the-art approach: (1) we introduce a new representation of co-location instances based on bounding boxes, (2) we formulate and prove several theorems regarding such a representation that can improve instances identification step, (3) we provide a novel algorithm utilizing the aforementioned theorems and analyze its complexity. Finally, we experimentally demonstrate the efficiency of the proposed solution.
05.09.2021
3 - 14
20
70