The following functionality was added or enhanced for version 2.6:
• Added native support for DOC, DOCX, PPT, PPTX, HTML and PDF.
• Added support for file metadata extraction.
• Added ISO 8601 compliant temporal normalization for absolute and relative time expressions.
• Enhanced ability for data scientists to add their own custom entity and entity attribute types.
• Enhanced Russian, Ukrainian, Belarusian, and Thai extraction capabilities.
• Enhanced Arabic, Russian, Spanish, and Thai lexicons.
• Enhanced logging.
About IMT Holdings
IMT Holdings, Corp. (IMT) is a next generation provider of multilingual natural language processing solutions. Through the Rosoka product line, multilingual tokenization, language identification, entity extraction, relationship extraction, entity salience and polarity (sentiment) analysis are all available for integration in a wide array of end-user products, from smart phones to enterprise data marts. More information on IMT Holdings and Rosoka can be found at www.imtholdings.com , www.rosoka.com, or @IMTHoldings on Twitter.
Multilingual Products for a Multilingual World.™