Trackback Spam Distinction Using Similarity of Contents
[Abstract]
An increase of trackback spam becomes a problem as blog
spreads, and it is necessary to be excluded. Distinction without people
is necessary because many of trackback spams are mechanically done.
Then, we experimented on the trackback spam distinction technique using
the similarity of the article. However, it is difficult to compare
content
of the article by the meaning. Then, it was considered that outline of
the
article was the same that the same noun as two articles was included.
As a result, rate of the track back spam rises when the similarity of
the
article is low.
-
藤村浩太、堀良彰、櫻井幸一、"内容の類似性を用いたトラックバックスパム判別”
情報処理学会 研究報告,2008年5月 pp19-24
-
Kohta Fujimura, Yoshiaki Hori, Kouichi Sakurai, "Trackback Spam
Distinction Using Similarity of Contents"
Joint Workshop on Information Security 2008