A co-location pattern is a set of spatial features frequently located together in space. A frequent pattern is a set of items that frequently appears in a transaction database. Since its introduction, the paradigm of frequent pattern mining has undergone a shift from candidate generation-and-test based approaches to projection based approaches. Co-location patterns resemble frequent patterns in many aspects. However, the lack of transaction concept, which is crucial in frequent pattern mining, makes the similar shift of paradigm in co-location pattern mining very difficult. This thesis investigates a projection based co-location pattern mining paradigm. In particular, a FP-tree based co-location mining framework and an algorithm called FP-CM, for FP-tree based co-location miner, are proposed. It is proved that FP-CM is complete, correct, and only requires a small constant number of database scans. The experimental results show that FP-CM outperforms candidate generation-and-test based co-location miner by an order of magnitude.
Identifer | oai:union.ndltd.org:unt.edu/info:ark/67531/metadc4724 |
Date | 05 1900 |
Creators | Yu, Ping |
Contributors | Huang, Yan, Mikler, Armin R., Brazile, Robert |
Publisher | University of North Texas |
Source Sets | University of North Texas |
Language | English |
Detected Language | English |
Type | Thesis or Dissertation |
Format | Text |
Rights | Public, Copyright, Yu, Ping, Copyright is held by the author, unless otherwise noted. All rights reserved. |
Page generated in 0.0018 seconds