Abstract
Surveillance is a critical measure to break anonymity. While surveillance with unlimited resources is often assumed as a means, against which, to design stronger anonymity algorithms, this paper addresses the general impact of limited resource on surveillance efficiency. The general impact of limited resource on identifying a hidden group is experimentally studied; the task of identification is only done by following communications between suspects, i.e., the information of whos talking to whom. The surveillance uses simple but intuitive algorithms to return more intelligence with limited resource. The surveillance subject used in this work is the publicly available Enron email data set, an actual trace of human interaction. The initial expectation was that, even with limited resource, intuitive surveillance algorithms would return the higher intelligence than a random approach by exploiting the general properties of power law-style communication map. To the contrary, the impact of limited resource was found large to the extent that intuitive algorithms do not return significantly higher intelligence than a random approach.
| Original language | English |
|---|---|
| Title of host publication | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
| Publisher | Springer Verlag |
| Pages | 81-99 |
| Number of pages | 19 |
| Volume | 8267 LNCS |
| ISBN (Print) | 9783319051482 |
| DOIs | |
| Publication status | Published - 2014 |
| Event | 14th International Workshop on Information Security Applications, WISA 2013 - Jeju Island, Korea, Republic of Duration: 2013 Aug 19 → 2013 Aug 21 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 8267 LNCS |
| ISSN (Print) | 03029743 |
| ISSN (Electronic) | 16113349 |
Other
| Other | 14th International Workshop on Information Security Applications, WISA 2013 |
|---|---|
| Country | Korea, Republic of |
| City | Jeju Island |
| Period | 13/8/19 → 13/8/21 |
Fingerprint
Keywords
- Anonymity
- Budget
- Email data set
- Surveillanc
ASJC Scopus subject areas
- Computer Science(all)
- Theoretical Computer Science
Cite this
Dynamic surveillance : A case study with Enron email data set. / Do, Heesung; Choi, Peter; Lee, Heejo.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8267 LNCS Springer Verlag, 2014. p. 81-99 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8267 LNCS).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Dynamic surveillance
T2 - A case study with Enron email data set
AU - Do, Heesung
AU - Choi, Peter
AU - Lee, Heejo
PY - 2014
Y1 - 2014
N2 - Surveillance is a critical measure to break anonymity. While surveillance with unlimited resources is often assumed as a means, against which, to design stronger anonymity algorithms, this paper addresses the general impact of limited resource on surveillance efficiency. The general impact of limited resource on identifying a hidden group is experimentally studied; the task of identification is only done by following communications between suspects, i.e., the information of whos talking to whom. The surveillance uses simple but intuitive algorithms to return more intelligence with limited resource. The surveillance subject used in this work is the publicly available Enron email data set, an actual trace of human interaction. The initial expectation was that, even with limited resource, intuitive surveillance algorithms would return the higher intelligence than a random approach by exploiting the general properties of power law-style communication map. To the contrary, the impact of limited resource was found large to the extent that intuitive algorithms do not return significantly higher intelligence than a random approach.
AB - Surveillance is a critical measure to break anonymity. While surveillance with unlimited resources is often assumed as a means, against which, to design stronger anonymity algorithms, this paper addresses the general impact of limited resource on surveillance efficiency. The general impact of limited resource on identifying a hidden group is experimentally studied; the task of identification is only done by following communications between suspects, i.e., the information of whos talking to whom. The surveillance uses simple but intuitive algorithms to return more intelligence with limited resource. The surveillance subject used in this work is the publicly available Enron email data set, an actual trace of human interaction. The initial expectation was that, even with limited resource, intuitive surveillance algorithms would return the higher intelligence than a random approach by exploiting the general properties of power law-style communication map. To the contrary, the impact of limited resource was found large to the extent that intuitive algorithms do not return significantly higher intelligence than a random approach.
KW - Anonymity
KW - Budget
KW - Email data set
KW - Surveillanc
UR - http://www.scopus.com/inward/record.url?scp=84958533915&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84958533915&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-05149-9_6
DO - 10.1007/978-3-319-05149-9_6
M3 - Conference contribution
AN - SCOPUS:84958533915
SN - 9783319051482
VL - 8267 LNCS
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 81
EP - 99
BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PB - Springer Verlag
ER -