TME 5.0: A feature-rich upgrade to Nstein's semantic solution

New linguistic features

Sentiment analysis is now faceted

Analyze the sentiment (tonality, subjectivity) not only at the document level, but also at the entity level. Determine what opinion is being expressed about a given person, place, trademark, and more.

Enhanced entity extraction

• Person Names cartridge now includes extended name typology (e.g. athletes, politicians, celebrities, etc.)
• Entities are now linked to a Unique Resource Identifier (URI) (e.g. Freebase, Wikipedia, etc.)
• New Trademarks/Products cartridge - now includes the ability to extract mentions of trademarks/products (i.e. any trademarks that refer to electronic devices - e.g. TVs, Cameras, MP3 players, Cell phones, GPS, etc. and car models - e.g. models from 2000 up 2010 & Classic cars). Extracted trademarks are categorized using the USPTO standard.
• New URLs cartridge - now includes the ability to extract URLs (i.e. any URLs that refer to a website, and email addresses)
• Extraction of locations embraces (hyper) local entities
• Extraction of locations includes geolocalisation attributes - available both in Coordinates (traditional sexagesimal conventions) and WGS-84 (World Geodetic System or GPS) formats
• Extraction of locations complies to the ISO3166-1 & ISO3166-2 standards
• Includes extended coverage of public/private companies, clubs, societies, associations, centers, foundations, orchestras, airports, laboratories, institutes, funds, universities, colleges, schools, government departments, ministries, organizations, resorts, museums, galleries, sport teams, music groups, etc.
• Includes stock tickers and ISO10383 standard Market Identifier Code (MIC) for the New York Stock Exchange, NASDAQ, Canadian TSX, London Stock Exchange and Euronext public companies

New TME Manager

Nstein's new TME Manager allows content management professionals to manage their taxonomies, thesauri, authority files, controlled lists, and knowledge bases by adding, deleting and editing terms using this web-based tool. Within this new TME Manager, all the user interfaces and underlying storage and workflow capabilities have been redesigned and optimized to make it easier than ever for internal users to manage their corporate knowledge structures. TME 5 places users at the center of metadata management.

New out-of-the-box, pre-trained, pre-packaged taxonomies

TME 5.0 comes out-of-the-box with version 19 of the IPTC taxonomy (in English, French and Spanish), the English ICB taxonomy (Industry Classification Benchmark), and the English Library of Congress thesaurus for graphic materials.

Major Spanish accuracy update

The overall Spanish concept, entity, and category extraction accuracy has been considerably improved.

Document type awareness

Nconcept Extractor is now asset aware, allowing speech segmentation and new out-of-the-box configurations for articles, photos, videos and audio assets.

New language detection module

Determines what language a given document is written in. Language detection automatically identifies the following: English, French, German, Arabic, Chinese (simplified), Czech, Danish, Estonian, Finnish, Hebrew, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Thai, Turkish, Ukrainian. Additional languages can be added upon request.

Updated technology

TME 5 is a next generation Text Mining Engine, and includes the latest technological developments to offer users a powerful set of tools to understand and enrich content efficiently.

RESTful distribution

REST (meaning Representational State Transfer) is a style of architecture widely considered to be the most open, seamless, scalable and interoperable set of protocols for any web application.

Enhanced performance

TME 5 has been reengineered to offer unprecedented performance and ensure that as content volume in today's world continues to increase.