Title: Center for Content Extraction

Release Date: 2014-06-18

Document Date: 2008-07-26

Description: This extract from a July 2008 NSA presentation explains the operation of the automatic name recognition tool NYMROD: see the Der Spiegel article ‘A’ for Angela: GCHQ and NSA Targeted Private German Companies and Merkel, 29 March 2013.

Document: TOP SECRET//COMINT//REL TO USA, AUS, CAN, GBR, NZUV20320108

Human Language
Technology

Center for Content Extraction

Chartered to distribute multi-lingual
Content Extraction Services to NS A
enterprise applications supporting
and enhancing 6 analytic functions:

Selection
T ranslation
Analysis

Investigative Research

Retrieval

Storage

T1221
July 2008

TOP SECRET//COMIMT//REL TO USA, AUS, CAN, GBR, NZL/V20320108

TOP SECRET/*OMINT//REL TO USA, AUS, CAN, GBR, NZU/20320108

r

NvmrodMissior

What analytic problems is Nymrod trying to
address?

- We need to search for the target behind what's been
extracted (find reported information about targeted
persons)

- HUMAN beings are central to most targets

- We need to cope with linguistic variation among entity
names (esp Person names)

- We need to resolve Entity "coreference" problems using
contextual information ("Evidence")

Human Language
Technology

TOP SECRET//COMINT//REL TO USA, AUS, CAN, GBR, NZU/20320108

TOP SECRET//#OMINT//REL TO USA, AUS, CAN, GBR, NZL//20320108

Top 10

Human Language
Technology

Most Frequent Mentions in Anchory Reports for the Previous Week

TOP SECRET//COMINT//REL TO USA, AUS, CAN, GBR, NZL//20320108

e-Highlighter

Click to send permalink to address bar, or right-click to copy permalink.

Un-highlight all Un-highlight selectionu Highlight selectionh