TNS Media Intelligence/Cymfony's Content Analysis Engine
TNS MI/Cymfony makes vast amounts of digital information useful to you and your organization. At TNS MI/Cymfony’s core is an advanced information extraction engine that combines information retrieval and Natural Language Processing (NLP) technologies to identify important people, places, companies, concepts, relationships and events in documents.
 |
| How our InfoXtract engine works |
Keyword-matching is not enough
While Internet search engines use a keyword-based search, our technology uses a more sophisticated form of information extraction based on a detailed grammatical analysis of the text.
TNS MI/Cymfony’s engine dissects article, paragraphs and sentences to determine who and what is being talked about, whether something or someone is a key focus or a passing reference, and how the various entities mentioned (companies, people, locations and events) relate to one another. Grammar-based approaches provide the meticulous detail necessary for useful information discovery, far more intelligent than keyword searching.
The engine combines grammar-based approaches with statistical models for natural language processing. Statistical models provide the robustness to rapidly adapt to new information sources. TNS MI/Cymfony has exploited the best of both paradigms by developing a unique, hybrid architecture that leverages both statistical and grammar-based approaches to deliver optimal speed and robustness.
Trusted by the U.S. Government
Initial funding for the development of the engine was provided through US Government research contracts and intelligence agencies. The Department of Defense has sponsored continued advancements in this technology over the last eight years in the areas of information extraction, machine learning and text analytics.
In 2005, TNS MI/Cymfony spun off its Government Systems business forming Janya Inc. TNS MI/Cymfony continues to focus on commercial enhancements to the InfoXtract engine while Janya focuses on government sponsored enhancements, particularly for DoD, intelligence and law enforcement communities.
|
