Nstein’s TME 5.0: Optimize Your Web Content for the Semantic Web

By David Roe (@druadh20) Jun 8, 2009

Nstein’s TME 5.0: Optimize Your Web Content for the Semantic Web With the planned Fall 2009 release of a feature-rich up-grade to its Text Mining Engine (TME), Nstein Technologies (news, site) is taking semantic metadata firmly into the world of Web 3.0.

For anyone that is not aware already, Nstein is a global specialist provider of solutions in the online media and web publishing world. TME 5.0, the company says, will include enhancements to linguistic abilities, Web 3.0 compliance and a number of new tools that will give users greater control over semantic metadata.

Next-Generation Capabilities

Now in its fifth iteration, TME is the engine that drives Nstein’s entire product range. Since its inception a decade ago, TME is the core intelligence driver to Nstein’s DAM (Digital Asset Management) and WCM (Web Content Management) solutions.

NsteinProducts[1].jpg
Text Mining Engine: at the heart of all Nstein solutions

With 12 years of development behind it — as well as considerable client services — TME 5.0 is RESTful, respects W3C standards and is fully Web 3.0 compatible.

“TME 5 delivers next-generation capabilities that allow large information providers to organize, share and discover information,” explained Jean-Michel Texier, CTO of Nstein. “More than ever we are finding our clients needing to rapidly innovate test and deploy new business strategies. TME 5 gives them unprecedented flexibility and control to support any business model …”

What this means is that TME 5.0 provides new taxonomies and comes with — literally — tens of thousands of names that have been pre-categorized as politicians, sports people, or celebrities.

Enhanced Linguistic Features

But that’s really the icing on the cake. Take a look under the bonnet and you find a high-octane engine that can parse sentences by identifying and extracting their grammatical structures.

Using enhanced semantic analysis it extracts what Nstein is describing as ‘linguistic DNA’ to enable searches that provide not only precise responses to search parameters, but also related content.

Nstein_TMEArchitecture[1].jpg
TME 5.0: Turns content into assets

What it does, in effect, is turn content into assets. By standardizing and adapting it, TME, by generating clean intelligent XML, will now create a content ‘asset’ that can be used for delivery across all digital channels.

Other enhanced linguistic features include:

Sentiment Analysis: An ability to analyze the sentiment of a document, as well as elements in that document and produce an analysis of the opinion, or sentiment, being expressed about a place, brand, person.
MetaData Aware: Content detection mode is now metadata aware allowing users to refine searches by user defined algorithms that can weigh different terms differently.
Language Aware: Automatic language detection for 21 languages with the ability to ad other languages on request.
Taxonomy Enhancements: Out of the box taxonomies with v19 of IPTC taxonomy, as well as Industry Classification Benchmark (ICB) and the Library of Congress thesaurus for graphic materials.

There is also a considerable improvement in its entity extraction with eight new features.