Complexity and Social Networks Blog: NYT magazine article: Can Network Theory Thwart Terrorsts?

February 2008
Sun	Mon	Tue	Wed	Thu	Fri	Sat
					1	2
3	4	5	6	7	8	9
10	11	12	13	14	15	16
17	18	19	20	21	22	23
24	25	26	27	28	29

« Network Visualization Tools | Main | The Social Affordances of Email in Japan and America »

14 March 2006

NYT magazine article: Can Network Theory Thwart Terrorsts?

There was a piece in the NYT Magazine on Sunday on the surveillance program, titled “Can Network Theory Thwart Terrorists?�? by Patrick Radden Keefe

It offers some basic ideas in applying network analysis to detecting terrorists, although it contains, I believe, a couple of mistakes. First, it states that “Network academics caution that the field is still in its infancy and should not be regarded as a panacea.�? Not a panacea, certainly true; but in its infancy? Certainly not—the modern field can trace its roots to the 1930s with the work of Moreno, Newcomb, and others. However, this is really just a quibble, in that the analysis of very large full network data (e.g., see Nathan Eagle’s posting on his work on the phone data of an entire country) is in its infancy, due to the lack of tools and data until recently.

A more serious apparent error is in the following statement:

“The use of such network-based analysis may explain the administration’s decision, shortly after 9/11, to circumvent the Foreign Surveillance Court. The court grants warrants on a case-by-case basis, authorizing comprehensive surveillance of specific individuals. The NSA program, which enjoys backdoor access to America’s major communication switches, appears to do just the opposite: the surveillance is typically much less intrusive than what a FISA warrant would permit, but it involves vast numbers of people.
In some ways, this is much less alarming than old-fashioned wiretapping. A computer that monitors the metadata of your phone calls and e-mail to see if you talk to terrorists will learn less about you than government agent listening in to the words you speak.�?

From what we have seen in the news reporting (excerpted in previous postings), however, the NSA program involves more than analysis of metadata. It also involves recording and content analysis through computer algorithms of the conversations of large numbers of individuals, with snippets selected out for human analysis. Indeed, I believe this was the whole basis of the controversy—the stories on the metadata came after the wiretapping story. While I wouldn’t say that metadata pose a non-intrusion, this is a much bigger deal. Indeed, as I believe I have said in previous postings, I would consider a digital recording of my conversation, content coded, to be a greater intrusion on my privacy than having two FBI agents listen to my conversations, with paper transcripts going into a file somewhere. Those bits do not decay like human memories, or easily lost, like files stored in the bowels of an enormous bureaucracy.

Posted by David Lazer at March 14, 2006 9:21 PM

Comments

Is anyone aware of any network theorists/practitioners that have made any connection between network research and the circumvention of FISA? Otherwise, any assertion of a connection to me seems specious.

After all, the administration's defense of its actions make reference to being able to be "nimble" and respond quickly. Collecting vast amounts of network data for analysis does not fit with this explanation.

Further, as such collection could presumably recieve authorization via a warrant, the fact that the data is network data is unrelated to the warrant controversy.

Perhaps the panacea of network analysis is in the flexibility afforded to reporters to make strong assertions without challenge.

Posted by: Brian Rubineau at March 15, 2006 11:18 AM

This reminds me of an older article that was going around a couple of years ago. I found a copy on USAToday's site:
http://www.usatoday.com/news/nation/2004-10-09-math-against-terror_x.htm?csp=36

It starts with a discussion of "order theory" in mathematics ( http://en.wikipedia.org/wiki/Order_theory ), but the last section of the article discusses the network work of fellow-traveler Kathleen Carley at CMU applied to revealing the dynamics of the leadership of Hamas (prior to that organization's democratic election). A quote from the article: "After Israel's assassination of Hamas founder Sheikh Ahmed Yassin in March, the program correctly predicted he would be succeeded by hard-liner Abdel Azziz Rantisi."

Posted by: Brian Rubineau at March 15, 2006 4:47 PM

A few quick clarifications. First, the collection of phone log data is not, to my knowledge, a violation of FISA-- which I believe only covers wiretapping/eavesdropping (anyone who knows better please post a comment). According to credible newspaper reports (see earlier postings), telecomms have been voluntarily sharing phone log data with the federal government. It would be surprising, if this information were being shared, that it was not being done in a timely (near real time) fashion. Second, phone companies have used simple kinds of link/pattern analysis (in part based on who calls whom) to identify certain likely cases of fraud-- and to respond to those cases rapidly. Telecomms have enormous computational power; the NSA has more. So, nimbleness is not a challenge here. Even Nathan Eagle (see earlier posting) has been able to analyze the phone log data of the entire UK. Nathan's talented, and has better computers than I do, but I think the NSA can bring an army more resources to the problem. Third, it is clear that the intelligence community has had an interest in network analysis.

It would therefore be surprising (and perhaps irresponsible) if, given (and having sought) all of these data, that the government didn't use simple analyses of phone log data to inform assessments as to who to monitor/investigate. Whether these data have been collected and managed in a legal and acceptable fashion is another question altogether. And whether this kind of analysis is particularly powerful, is another (empirical) question-- one that maybe 100 years from now the burgeoning field of social network archival historians will find the answer to. My own intuition is that there are some weak but useful signals to distill in the most obvious analyses to conduct, but nothing more-- to those historians in 2106, please cite this blog. ; )

Posted by: David Lazer at March 15, 2006 10:35 PM

Complexity and Social Networks Blog

February 2008

4

5

6

8

9

19

24

25

26

28

Editor Login

Convener in chief:

Editors:

Categories

Archives

Recent Entries

Recent Comments

Notification

NYT magazine article: Can Network Theory Thwart Terrorsts?

Comments

Notification

Post a comment