The link structure of the Web is generally viewed as a webgraph. One of the main objectives of web structure mining is to find hidden communities on the Web based on the webgraph, and one of its approaches tries to enumerate substructures, each of which corresponds to a set of web pages of a community or its core. Research has shown that certain substructures can find sets of pages that are inherently irrelevant to communities. In this paper, we propose a model, which we call contracted webgraphs, where such substructures are contracted into single nodes to hide useless information. We then try structure mining iteratively on those contracted webgraphs since we can expect to find further hidden information once irrelevant information is eliminated. We also explore the structural properties of contracted webgraphs from the viewpoint of scale-freeness, and we observe that they exhibit novel and extreme self-similarities.
Yushi UNO
Osaka Prefecture University
Fumiya OGURI
Nihon Software Corporation, Ltd.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Yushi UNO, Fumiya OGURI, "Contracted Webgraphs — Scale-Freeness and Structure Mining —" in IEICE TRANSACTIONS on Communications,
vol. E96-B, no. 11, pp. 2766-2773, November 2013, doi: 10.1587/transcom.E96.B.2766.
Abstract: The link structure of the Web is generally viewed as a webgraph. One of the main objectives of web structure mining is to find hidden communities on the Web based on the webgraph, and one of its approaches tries to enumerate substructures, each of which corresponds to a set of web pages of a community or its core. Research has shown that certain substructures can find sets of pages that are inherently irrelevant to communities. In this paper, we propose a model, which we call contracted webgraphs, where such substructures are contracted into single nodes to hide useless information. We then try structure mining iteratively on those contracted webgraphs since we can expect to find further hidden information once irrelevant information is eliminated. We also explore the structural properties of contracted webgraphs from the viewpoint of scale-freeness, and we observe that they exhibit novel and extreme self-similarities.
URL: https://global.ieice.org/en_transactions/communications/10.1587/transcom.E96.B.2766/_p
Copy
@ARTICLE{e96-b_11_2766,
author={Yushi UNO, Fumiya OGURI, },
journal={IEICE TRANSACTIONS on Communications},
title={Contracted Webgraphs — Scale-Freeness and Structure Mining —},
year={2013},
volume={E96-B},
number={11},
pages={2766-2773},
abstract={The link structure of the Web is generally viewed as a webgraph. One of the main objectives of web structure mining is to find hidden communities on the Web based on the webgraph, and one of its approaches tries to enumerate substructures, each of which corresponds to a set of web pages of a community or its core. Research has shown that certain substructures can find sets of pages that are inherently irrelevant to communities. In this paper, we propose a model, which we call contracted webgraphs, where such substructures are contracted into single nodes to hide useless information. We then try structure mining iteratively on those contracted webgraphs since we can expect to find further hidden information once irrelevant information is eliminated. We also explore the structural properties of contracted webgraphs from the viewpoint of scale-freeness, and we observe that they exhibit novel and extreme self-similarities.},
keywords={},
doi={10.1587/transcom.E96.B.2766},
ISSN={1745-1345},
month={November},}
Copy
TY - JOUR
TI - Contracted Webgraphs — Scale-Freeness and Structure Mining —
T2 - IEICE TRANSACTIONS on Communications
SP - 2766
EP - 2773
AU - Yushi UNO
AU - Fumiya OGURI
PY - 2013
DO - 10.1587/transcom.E96.B.2766
JO - IEICE TRANSACTIONS on Communications
SN - 1745-1345
VL - E96-B
IS - 11
JA - IEICE TRANSACTIONS on Communications
Y1 - November 2013
AB - The link structure of the Web is generally viewed as a webgraph. One of the main objectives of web structure mining is to find hidden communities on the Web based on the webgraph, and one of its approaches tries to enumerate substructures, each of which corresponds to a set of web pages of a community or its core. Research has shown that certain substructures can find sets of pages that are inherently irrelevant to communities. In this paper, we propose a model, which we call contracted webgraphs, where such substructures are contracted into single nodes to hide useless information. We then try structure mining iteratively on those contracted webgraphs since we can expect to find further hidden information once irrelevant information is eliminated. We also explore the structural properties of contracted webgraphs from the viewpoint of scale-freeness, and we observe that they exhibit novel and extreme self-similarities.
ER -