A system and method are disclosed for determining the transcription/translation quality of web video content based on metrics derived from indirect feedback of users on existing captioning. The method may take into account how often the closed caption (CC) option was activated by a user on videos and the number of times users stayed through the whole video content using the closed captions. The system can also be used in assessing quality of manual transcription for languages that do not have automated speech recognition and to validate acoustic and language models in machine translation/transcription.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Kanevsky, Dimitri and Mao, Mark, "Method For Providing Metrics To Determine Transcription/Translation Quality", Technical Disclosure Commons, (November 30, 2015)