WWW Cache Layout to Ease Network Overload

Kenichi Yoshida
Proceedings of the Sixth International Workshop on Artificial Intelligence and Statistics, PMLR R1:537-548, 1997.

Abstract

The GBI (graph-based induction) concept learning method is applied to extract typical access patterns of WWW data. By interpreting extracted patterns as the cache site layout we can reduce the total network data flow by implementing a distributed cache system which is adapted to the WWW access patterns. Although the huge WWW data flow causes the overflow of the conventional hierarchical cache system, the layout created by the GBI method eases this problem. The traffic reduction ratio of this distributed cache system is 2.5 times higher than that of the conventional hierarchical cache system. Our results suggest the importance of the data analyzing methods which can handle structured data. By analyzing regularity in graph structures, the GBI method can reduce the network data flow; The statistical criteria contribute to the analysis of promising patterns

Cite this Paper

BibTeX
@InProceedings{pmlr-vR1-yoshida97a, title = {WWW Cache Layout to Ease Network Overload}, author = {Yoshida, Kenichi}, booktitle = {Proceedings of the Sixth International Workshop on Artificial Intelligence and Statistics}, pages = {537--548}, year = {1997}, editor = {Madigan, David and Smyth, Padhraic}, volume = {R1}, series = {Proceedings of Machine Learning Research}, month = {04--07 Jan}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/r1/yoshida97a/yoshida97a.pdf}, url = {https://proceedings.mlr.press/r1/yoshida97a.html}, abstract = {The GBI (graph-based induction) concept learning method is applied to extract typical access patterns of WWW data. By interpreting extracted patterns as the cache site layout we can reduce the total network data flow by implementing a distributed cache system which is adapted to the WWW access patterns. Although the huge WWW data flow causes the overflow of the conventional hierarchical cache system, the layout created by the GBI method eases this problem. The traffic reduction ratio of this distributed cache system is 2.5 times higher than that of the conventional hierarchical cache system. Our results suggest the importance of the data analyzing methods which can handle structured data. By analyzing regularity in graph structures, the GBI method can reduce the network data flow; The statistical criteria contribute to the analysis of promising patterns}, note = {Reissued by PMLR on 30 March 2021.} }
Endnote
%0 Conference Paper %T WWW Cache Layout to Ease Network Overload %A Kenichi Yoshida %B Proceedings of the Sixth International Workshop on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 1997 %E David Madigan %E Padhraic Smyth %F pmlr-vR1-yoshida97a %I PMLR %P 537--548 %U https://proceedings.mlr.press/r1/yoshida97a.html %V R1 %X The GBI (graph-based induction) concept learning method is applied to extract typical access patterns of WWW data. By interpreting extracted patterns as the cache site layout we can reduce the total network data flow by implementing a distributed cache system which is adapted to the WWW access patterns. Although the huge WWW data flow causes the overflow of the conventional hierarchical cache system, the layout created by the GBI method eases this problem. The traffic reduction ratio of this distributed cache system is 2.5 times higher than that of the conventional hierarchical cache system. Our results suggest the importance of the data analyzing methods which can handle structured data. By analyzing regularity in graph structures, the GBI method can reduce the network data flow; The statistical criteria contribute to the analysis of promising patterns %Z Reissued by PMLR on 30 March 2021.
APA
Yoshida, K.. (1997). WWW Cache Layout to Ease Network Overload. Proceedings of the Sixth International Workshop on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research R1:537-548 Available from https://proceedings.mlr.press/r1/yoshida97a.html. Reissued by PMLR on 30 March 2021.

Related Material