According to the present invention, a speech to text (STT) translation method for generating a subtitle of a video in a server using dialect DB to convert a dialect voice signal into a standard language subtitle and provide the standard language subtitle may comprise the steps of: constructing database in which dialect texts and standard language texts corresponding to the dialect texts match; obtaining an STT result in response to a subtitle provision request for target video; providing the STT result as a subtitle for target video if a specific dialect text stored in the database is not included in the STT resu and converting the specific dialect text into a specific standard language text corresponding to the specific dialect text to provide the converted specific standard language text as a subtitle for target video from the STT result if the dialect text stored in the database is included in the STT result.
展开▼