Presentation 2008-12-17
A Masquerade Detecting Method Based On the TF-IDF Model
Geng Dai, Zhou Jian, Haruhiko Shirai, Jousuke Kuroiwa, Tomohire Odaka, Hisakazu Ogura,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Masquerade is someone who impersonates another user and operates computer system with privileged access. Computer security caused by masquerade is in a serious situation. Many researches have been proposed from different viewpoints. For the variability of user behavior and the similarity between the behavior of masquerade and that of normal user, it's difficult to gain a satisfy detection efficiency. Especially, these researches conducted their experiments on the benchmark datasets of Unix command sequence, so that efficiency of different methods could be compared. In this study, we proposed to apply the Term Frequency/Inverse Document Frequency (TF-IDF) model, which has been widely used in text classification in Natural Language Processing (NLP), to masquerade detection. Command was treated as word, and session was treated as paragraph. Therefore, the classification problem between normal and masquerade could be looked as the classification of text. A new formula of TF-IDF was proposed, and a ranking scheme was also further presented to improve detection efficiency. Experiment was conducted on a benchmark data, and experimental results show that it reports better detection efficiency than that of the previous Uniqueness method. Though, only the characteristic of command frequency is considered at the moment, it gains an encouraging detection efficiency. A better efficiency is expected to be gained in future research by updating.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Masquerade detection / Command sequence / TF-IDF / Unix
Paper # ISEC2008-94
Date of Issue

Conference Information
Committee ISEC
Conference Date 2008/12/10(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Information Security (ISEC)
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Masquerade Detecting Method Based On the TF-IDF Model
Sub Title (in English)
Keyword(1) Masquerade detection
Keyword(2) Command sequence
Keyword(3) TF-IDF
Keyword(4) Unix
1st Author's Name Geng Dai
1st Author's Affiliation University of Fukui()
2nd Author's Name Zhou Jian
2nd Author's Affiliation Univ. of Fukui
3rd Author's Name Haruhiko Shirai
3rd Author's Affiliation Univ. of Fukui
4th Author's Name Jousuke Kuroiwa
4th Author's Affiliation Univ. of Fukui
5th Author's Name Tomohire Odaka
5th Author's Affiliation Univ. of Fukui
6th Author's Name Hisakazu Ogura
6th Author's Affiliation Univ. of Fukui
Date 2008-12-17
Paper # ISEC2008-94
Volume (vol) vol.108
Number (no) 355
Page pp.pp.-
#Pages 6
Date of Issue