Mining and filtering multi-level spatial association rules with ARES
ISMIS'05 Proceedings of the 15th international conference on Foundations of Intelligent Systems
| Hi-index | 0.00 |
The tremendous growth in data has generated the need for new techniques that can intelligently transform the massive data into useful information and knowledge. Data mining is such a technique that extracts nontrivial, implicit, previously unknown, and potentially useful information from data in databases. Association rule mining is one of the important advances in data mining. In this dissertation, we propose a comprehensive model to derive association rules on Remote Sensed Imagery (RSI) data. Discovery of interesting patterns and rules from RSI datasets composed of images and associated ground data can be of importance in precision agriculture and other areas. However, the image data sizes are usually too large to be mined in a reasonable amount of time with existing algorithms. In this dissertation, we use a lossless and compressed structure, called Peano Count Tree (P-tree), to derive association rules on RSI data. Based on P-trees, an efficient association rule mining algorithm, P-ARM, with fast support calculation and significant pruning techniques was proposed. Experiments show that the P-ARM algorithm outperforms other algorithms, such as Apriori and FP-growth. We also introduce an approach to derive high confident rules efficiently using a data cube called Tuple Count Cube (T-cube). In addition, we present a framework for parallel association rule mining on RSI data. In this framework, by using active network and P-trees, both the communication cost and the local execution time are reduced.