Name
Abstract | ||
|---|---|---|
The processing of a Continuous Reverse k-Nearest-Neighbor (CRkNN) query on moving objects can be divided into two sub tasks: continuous filter, and continuous refinement. The algorithms for the two tasks can be completely independent. Existing CRkNN solutions employ Continuous k-Nearest-Neighbor (CkNN) queries for both continuous filter and continuous refinement. We analyze the CkNN based solution and point out that when k1 the refinement cost becomes the system bottleneck. We propose a new continuous refinement method called CRange-k. In CRange-k, we transform the continuous verification problem into a Continuous Range-k query, which is also defined in this paper, and process it efficiently. Experimental study shows that the CRkNN solution based on our CRange-k refinement method is more efficient and scalable than the state-of-the-art CRkNN solution. |
Year | DOI | Venue |
|---|---|---|
2008 | 10.1109/MDM.2008.31 | MDM |
Keywords | Field | DocType |
crange-k refinement method,new continuous refinement method,continuous filter,continuous k-nearest-neighbor,refinement cost,continuous refinement,crknn solution,continuous range-k query,state-of-the-art crknn solution,k-nearest-neighbor monitoring,continuous verification problem,application software,uncertainty,filtering,game theory,mobile computing,algorithm design and analysis,mobile communication,pattern recognition,k nearest neighbor | k-nearest neighbors algorithm,Bottleneck,Algorithm design,Pattern recognition,Computer science,Algorithm,Filter (signal processing),Artificial intelligence,Mobile telephony,Distributed computing,Scalability | Conference |
Citations | PageRank | References |
31 | 1.09 | 14 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Wei Wu | 1 | 160 | 13.27 |
| Fei Yang | 2 | 31 | 1.09 |
| Chee Yong Chan | 3 | 643 | 199.24 |
| Kian-Lee Tan | 4 | 6962 | 776.65 |