This dataset was recorded in a quiet office environment, with the participation of 20 speakers, including 10 males and 10 females. All speakers involved in the recording were professionally selected to ensure standardized pronunciation and clear enunciation. The recorded texts cover live broadcast comments and other information.