Products / Text Mining Engine / TME Features

Increase ROI by Leveraging the 'Aboutness' of Content

Built to process high volumes of multilingual, unstructured information, Nstein's Text Mining Engine complies with press and media industry requirements for ongoing content production workflows, automatically generating contextual metatags from sources such as newsroom articles, newsfeeds, audio and video files, blogs, forums and user-generated content.

Nstein's TME also provides a highly productive content management environment that offers benefits to editorial staff and knowledge workers. Going beyond mere keyword-based technology, it delivers faster and more accurate search results, seamlessly integrates with existing workflows and enables highly profitable categorization and repurposing of both live content and archives.

Available Modules
Nconcept Extractor
  • Intelligently locates and extracts key concepts (keywords and noun phrases) within any document, creating a unique snapshot of a document's most important ideas
  • Automates extraction and ranking of concepts to isolate subjects and themes within content
  • Combines linguistics-based and statistics-backed processes with great sophistication
  • Patented technology (LDNA™-Linguistic DNA)
Ncategorizer
  • Automatically indexes or classifies documents into categories / terms, and shows category relevancies
  • Takes extracted concepts from documents and maps non-controlled terms (i.e. the concepts) to a controlled vocabulary (e.g. a customer's) that may include taxonomy, thesaurus, ontology, authority file or other lists
  • Ncategorizer comes trained on the IPTC taxonomy out-of-the-box (available only for some languages)
Nfinder
  • Automates extraction of entities (controlled terms). Relies on machine learning and linguistic rules and works with an organization's authority files and dictionaries as well as with Nstein's pre-built authority files
  • Can be customized with existing client authority files
  • Some of the entity cartridges available:
    • People
    • Geographical locations (country, city, state, province, etc.)
    • Organizations (company names, etc.)

Additional add-on text mining packages also available: Nsummarizer (document summaries), Nserver Manager (back office with Authority Files & Taxonomy Manager, and Knowledge Base Manager), Nsentiment (tone detector), Nlikethis (similar documents) and Nretriever (topic clustering).

For a complete list of modules and features, please download Nstein's TME Data Sheet.

TME Architecture Diagram

TME Data Sheet

Download TME's features, benefits and specs.

TME Specifications

Server Platform:

  • Dual ®Xeon, 3.6 GHz+
  • 4GB RAM
  • 146GB+ Ultra Wide HD

OS:

  • Linux Red Hat 4 & 5
  • Windows 2003

Programming Interfaces:

  • XML-RPC (socket), JAVA, .NET, C++, PHP connectors, Web Services (SOAP)

White Paper - Nsentiment

Get the pulse of your readers by analyzing their sentiment on blogs, forums and user-generated content.