Приложение на ASR на български език за публикуване на преки репортерски включвания в интернет издания
ASR App in the Bulgarian Language for Publishing Live Reporter Inclusions in Online Editions

Author(s): Ventsislav Vassilev
Subject(s): Social Sciences, Media studies, Communication studies
Published by: Факултет по журналистика и масова комуникация, Софийски университет „Св. Кл. Охридски”
Keywords: online journalism; ASR, AI in media; Natural language processing (NLP); Speech-to-text (STT); voice recognition
Summary/Abstract: The following research studies current Automatic speech recognition (ASR) software in the Bulgarian language and its practical use in media production for decrypting live broadcasts. Several live broadcasts from Bulgarian National Radio are transcribed by AI and humans alike. After that, professional editors compared both texts according to several criteria – time to decrypt; accuracy in grammar, and word recognition. A comparison was made about what time it takes a professional web editor to adapt human and AI-generated text. Yet another comparison is made to determine if using AI decryptions and not human labor does indeed save time when these live reports are to be published on a media website.

• Details
• Contents