SEARCH THE RADIO (Round 4)

Summary
The solution
The project is based on the key observation that valuable audio content is digitally inaccessible and difficult to search. The aim is to use existing automatic audio-to-text transcription technology in a smarter way – turning the most valuable Radio TOK audio content into fully searchable and accessible text for listeners and every internet user. The database grows daily by approximately 20 hours of unique talk audio content. By establishing a process of asynchronous transcription of newly produced audio content and additionally transcribing the most valuable pieces of the archives, a database of carefully indexed transcriptions with intelligent search ability will be produced.