by Rachel Ehrenfeld May 14th 2011
Talking on the CBS show “60 Minutes,” President Obama noted: “It’s going to take some time for us to exploit the intelligence that we were able to gather on site” during the raid in which Osama bin Laden was killed. This information, according to Mr. Obama, gives the U.S. a chance “to…really deliver a fatal blow [to Al Qaeda], if we follow through aggressively in the months to come.”
The United States intelligence agencies reliance on labor-intensive, time-consuming and inefficient methods to decipher the captured electronic devices and paper documents. The delay has allowed al Qaeda operatives and many of its budding affiliates to relocate, change their identity and communication methods, diminishing the U.S. ability to act upon the intelligence contain in the trove, The window of opportunity to destroy the organization and its global metastasizes, has significantly shrunk.
This event demonstrates the need for a more efficient process to mine the tens of millions of documents – electronic and print – which are seized or intercepted annually in different languages by U.S. intelligence, military, and law enforcement agencies. There is an urgent need to extract information of high intelligence value contained in such documents rapidly and efficiently, and to pass the intelligence on for operational exploitation.
The current Document Exploitation (DOCEX) method, including automated translation and categorization technologies, suffers from several major deficiencies: focus on an inefficient key word spotting techniques, long processing times, and a high rate of false negatives and positives. Moreover, the automated translation tools are still unsuitable for the sophisticated language of neo-classical Arabic used by Islamist terrorists such as al-Qaeda. Similarly, categorization technologies do not meet the challenge of analyzing sophisticated, culturally sensitive nuances. Poorly translated and categorized documents of course cause the loss of crucial actionable and preventative intelligence.
To address such problems, IntuView, an Israeli company, has developed revolutionary software that can process large volumes of digital documents related to Islamic terrorism very quickly. The software instantly assesses any digital document in a supported language (Arabic, Urdu, Farsi, etc.), determines its relevance and risk rating, and provides an intelligence analysis report based, for example, on the content of the document, including classification, date of writing, type of document, author, region, ideological, ethnical and religious affiliation, and a summary of the content, based on a sophisticated “artificial intuition” program. The summary integrates the implicit meaning of religious concepts, verses from religious and ideological texts, and historical allusions. The software integrates the information about entities (individuals, organizations, etc.) mentioned in the documents to create a virtual “identity card” of the person behind that name: possible affiliations, name variants, ethnic origin, gender, family/tribal links, etc., aggregating and matching information found in multiple entries and identifying links (family, tribe) between the entities.
This software implements linguistic programs that look for entities in the texts that could be “ideas”, “actions”, “persons” “groups,” etc. An idea may be composed of statements in different parts of the document, which are revealed only when combined. This “idea” is tagged with a unique meaning (as opposed to words which may have different meanings in different contexts). The user then receives an executive summary in English detailing what the document is about, who wrote it, when it was written, who it is for, who it is directed against, etc.
Allusions from cultural or religiously oriented quotes, such verses from the Qur’an which “mean” in al-Qaeda documents permission for attacks, are extracted. And a sophisticated program aggregates information on individuals mentioned in the documents matching them with the database of the user. If categorization, translation and analysis of 1000 documents take over 200 man-days using normal methods, the same process takes less than 2 hours with this system. Some U.S. military, border control and law enforcement agencies are already using this software, but that process has yet to be fully utilized. The sooner this software can be made available and widely applied, the better for our national security.
A variant dedicated to terrorist financing by legal and illegal entities in now being developed in cooperation with the New York based American Center for Democracy (ACD). The financial intelligence software will help expose and prevent the use of legal and laundered money to penetrate and undermine the U.S. economy, and enhance the intelligence capabilities of the military, law enforcement and business communities.
Bio: Dr. Rachel Ehrenfeld is the director of the ACD. She is the author of Funding Evil: How Terrorism is Financed – and How to Stop It.