Comment by xdfgh1112

I suspect podcasts, as you have a huge amount of transcribed data with good diction and mic quality. The voices sound like podcast voices to me.