Name
Abstract | ||
|---|---|---|
In USN system, data streams are generated from several sensor nodes continuously. Therefore, patterns from data streams can be occur continuously and be a very long sequence pattern. However, it is very expensive to mine a very long sequence pattern using existing pattern mining methods. In this paper, we define the Pattern Map to store serial patterns and we propose the Pattern Map mining to discover useful information from the Pattern Map in real-time. Also, we define the Meaning Window to produce transactions from data streams and we define the HAPM(HAsh based Pattern Map) as a new data structure to represent the Pattern Map. Finally, we introduce Local Rules and the Global Rules as a new rule concept for USN application environments. |
Year | DOI | Venue |
|---|---|---|
2012 | 10.1145/2184751.2184843 | ICUIMC |
Keywords | Field | DocType |
long sequence pattern,pattern map mining,global rules,pattern map,new data structure,real-time pattern map mining,serial pattern,usn application environment,data stream,local rules,existing pattern mining method,data structure,real time | Data mining,Data structure,Data stream mining,Data stream,Computer science,Hash function,Sequence pattern | Conference |
Citations | PageRank | References |
1 | 0.41 | 6 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Jae-In Kim | 1 | 8 | 2.42 |
| Pilsun Choi | 2 | 5 | 1.65 |
| Buhyun Hwang | 3 | 112 | 24.15 |