This dataset was recorded in a quiet office environment, with the participation of 1,682 speakers, including 692 males and 990 females. All speakers involved in the recording were professionally selected to ensure standardized pronunciation and clear enunciation. The recorded texts cover daily spoken language, education, medical health, sports and entertainment, and social economy.