Ping-Pong is an algorithm which outputs a certain type of dense submatrices from a sparse binary matrix. In this paper we extend the algorithm and solve its drawbacks. To make extensions, we review the algorithm in two different ways. First, we see it as a graph-search, make its state-space explicit, and introduce another search method than that of the original Ping-Pong. Secondly, we see it as decision of similar rows and columns on a binary matrix, and introduce another similarity measure than that of the original Ping-Pong. By use of a text dataset, we CONFIRM that these extensions resolve original drawbacks
Description
Science Links Japan | Extensions of Ping-Pong Algorithm
%0 Generic
%1 kamiharako2006extensions
%A Kamiharako, M.
%A Oyanagi, S.
%B IPSJ Transactions on Database
%D 2006
%K 2009 clustering co-clustering extension ping-pong seminar
%P 1-15
%T Extensions of Ping-Pong Algorithm
%U http://sciencelinks.jp/j-east/article/200620/000020062006A0742991.php
%X Ping-Pong is an algorithm which outputs a certain type of dense submatrices from a sparse binary matrix. In this paper we extend the algorithm and solve its drawbacks. To make extensions, we review the algorithm in two different ways. First, we see it as a graph-search, make its state-space explicit, and introduce another search method than that of the original Ping-Pong. Secondly, we see it as decision of similar rows and columns on a binary matrix, and introduce another similarity measure than that of the original Ping-Pong. By use of a text dataset, we CONFIRM that these extensions resolve original drawbacks
@misc{kamiharako2006extensions,
abstract = {Ping-Pong is an algorithm which outputs a certain type of dense submatrices from a sparse binary matrix. In this paper we extend the algorithm and solve its drawbacks. To make extensions, we review the algorithm in two different ways. First, we see it as a graph-search, make its state-space explicit, and introduce another search method than that of the original Ping-Pong. Secondly, we see it as decision of similar rows and columns on a binary matrix, and introduce another similarity measure than that of the original Ping-Pong. By use of a text dataset, we CONFIRM that these extensions resolve original drawbacks},
added-at = {2009-12-14T13:05:38.000+0100},
author = {Kamiharako, M. and Oyanagi, S.},
biburl = {https://www.bibsonomy.org/bibtex/2dd4d115132429cc6d37f5d46a7100d09/r.b.},
booktitle = {IPSJ Transactions on Database},
description = {Science Links Japan | Extensions of Ping-Pong Algorithm},
interhash = {9c6534e6fe2740055bf4b2c7b8af4ba0},
intrahash = {dd4d115132429cc6d37f5d46a7100d09},
keywords = {2009 clustering co-clustering extension ping-pong seminar},
pages = {1-15},
timestamp = {2009-12-14T13:05:38.000+0100},
title = {Extensions of Ping-Pong Algorithm},
url = {http://sciencelinks.jp/j-east/article/200620/000020062006A0742991.php},
year = 2006
}