Abstract
With the development of networks, the distributed data processing is increasingly becoming important. A distributed clustering system is a typical distributed data mining, and is an important part for distributed information retrieval. Based on the multi-agent environment (MAGE), the distributed clustering system will be developed, which takes the growing hierarchical self-organizing map (GHSOM) as the main learning algorithm. The MAGE system offers a well platform for acquiring information and sharing knowledge in the distributed environment. The dynamic growing architecture of GHSOM well matches the deficiency condition of lacking the complete information in distributed environment; the hierarchical architecture organizes the clusters of data better; also the resulting map can serve as retrieval interfaces. This distributed clustering system can be well utilized in distributed information retrieval systems such as a distributed image retrieval system.
| Original language | English |
|---|---|
| Pages (from-to) | 346-350 |
| Number of pages | 5 |
| Journal | Harbin Gongcheng Daxue Xuebao/Journal of Harbin Engineering University |
| Volume | 27 |
| Issue number | SUPPL. |
| State | Published - Jul 2006 |
| Externally published | Yes |
Keywords
- Agent
- Clustering
- Distributed system
- GHSOM