April 14, 2009
AppTek’s MediaSphere seamlessly integrates hybrid machine translation, speech recognition, name entity identification and multilingual information retrieval
Published: April 14, 2009
McLean, VA (April 14, 2009) – AppTek, a leader in human language technology (HLT), is announcing key enhancements to its media monitoring solution, MediaSphereTM. A software solution providing multilingual transcripts of various television and radio stations for many domestic and international news bureaus as well as transcripts for telephony, MediaSphere now incorporates the company’s hybrid machine translation (HMT) system as well as increased coverage of dialects for automated speech recognition (ASR) and new engines for non- broadcast sources. AppTek is demonstrating the new capabilities of MediaSphere later today during the Defense Intelligence Agency’s Human Language Technology (HLT) Showcase in Washington, D.C. (www.ncsi.com/diahlts09/exhibiting_opps.shtml).
MediaSphere is a turnkey solution that utilizes AppTek’s speaker adaptive speech recognition engine to adjust to changes in dialect and language in real-time to improve speed and accuracy of all media channel monitoring. The updated media monitoring software seamlessly integrates AppTek’s HMT system with its ASR engine to provide a unified and scalable solution.
MediaSphere offers text processing with advanced linguistic capabilities such as speech-to-text transcription, machine translation of transcribed text, information retrieval with query translation and automated name-entity detection. These enhancements enable MediaSphere to generate transcripts from both broadcast and non-broadcast sources, translate the transcribed text and deliver rich media content online.
“AppTek is the only provider with full hybrid machine translation, using both rule-based and statistical machine translation for better translation fluency, accuracy and informativeness,” said Steve Cook, CTO at AppTek. “Because AppTek owns all of the necessary technologies, we are uniquely positioned to provide our customers with a complete out-of-the-box solution that seamlessly integrates all aspects of media monitoring from speech recognition and translation to the delivery of meaningful and actionable information.”
MediaSphere provides 24/7 monitoring of live channels in a variety of languages in a fully- automated system that provides alerts and information based on customized reporting and search parameters set by the user. Utilizing video and audio logging technologies, transcribed content is time-coded, analyzed and indexed, allowing the user to query the repository for terms of interest. Matching terms are highlighted within the context in which they were mentioned. Advanced analysis technologies, including name-entity detection, clustering and categorization, allow the user to perform intelligent content and meta-data queries.