Cyber Security

DARPA developing ultimate web search engine to police the internet

  Published time: February 12, 2014 22:35                                                                            

Reuters / Umit BektasReuters / Umit Bektas

​The Pentagon’s research arm that fosters futuristic technology for the military will soon begin working to surpass current abilities of commercial web search engines. Yet, once it masters the “deep Web,” the agency doesn’t say much about what comes next.

The Defense Advanced Research (DARPA) said its “Memex”   project will be able to search the far corners of internet  content that is unattainable by modern, mainstream search  engines, offering DARPA “technological superiority in the  area of content indexing and Web search on the Internet.”

DARPA said earlier this month in its solicitation announcement  for Memex proposals that the system will initially be used to  counter human trafficking, which often thrives in web forums,  chat rooms, job postings, hidden services and other websites.

To root out trafficking operations within the invisible corners  of the web, commonly referred to as the “deep web,”   Memex (a melding of “memory” and “index”)   “will address the inherent shortcomings of centralized search  by developing technology for domain-specific indexing of Web  content and domain-specific search capabilities.”

With Memex, DARPA hopes to achieve the ability for decentralized,  automated, topic-precise searches that can leverage image  recognition and natural language technology.

DARPA has asked researchers to develop advanced web-crawler  software to reach sites and resources that have sophisticated  crawler defenses. Memex operators would then be able to access  the indexed domain-relevant content with much greater precision  and ease than is currently possible.

Memex, DARPA says, will be first employed against human  trafficking, which, “especially for the commercial sex trade,  is a line of business with significant Web presence to attract  customers and is relevant to many types of military, law  enforcement, and intelligence investigations.”

DARPA says that dark places online where trafficking occurs  enables “a growing industry of modern slavery” that can  be stopped with Memex capabilities.

“An index curated for the counter trafficking domain,  including labor and sex trafficking, along with configurable  interfaces for search and analysis will enable a new opportunity  for military, law enforcement, legal, and intelligence actions to  be taken against trafficking enterprises,” DARPA’s  solicitation announcement reads.

Yet while DARPA mentions the usefulness of such technology for  law enforcement and investigative purposes regarding human  trafficking – basically, crimes few are opposed to stopping – it  does not address the myriad other uses Memex would offer the US  military, government intelligence operations, or police actions.

Amid the recent disclosures of government spying via the National  Security Agency’s operations, the topic of complete surveillance  over the entirety of the web is a sore subject. Thus, DARPA says  it is “specifically not interested in proposals for the  following: attributing anonymous services deanonymizing or  attributing identity to servers or IP addresses, or gaining  access to information which is not intended to be publicly  available.”

How DARPA would catch traffickers without   “deanonymizing” someone, though, the agency does not  explain. Nor does it address just how far it wants to out anyone  hiding in the deep web for legitimate reasons, whether they are  journalists, whistleblowers, activists, and the like.

The Memex project takes its name from a 1945 article in The  Atlantic titled “As We May Think,” by Dr. Vannevar Bush,  head of the White House Office of Scientific Research and  Development. Bush envisioned a “device” that could be  used for finding and categorizing the world’s information, acting  as a supplement for the human brain.

“In a nutshell, Bush wanted to mimic how the human brain  thinks, learns, and remembers information,” writes  Motherboard. “Which is exactly what artificial intelligence  researchers at the DoD and in Silicon Valley are trying to do  now, to glean better insights from the unruly army of big data  being collected by web giants and the military alike.”

The Memex project is expected to run over the next three years,  with proposals due in April

http://rt.com/usa/darpa-internet-search-engine-788/

Enhanced by Zemanta