how do you populate/maintain the synonym/stemming database? is this something that “standard” datasets exist for? asking because this problem will fall on my plate at work sometimes in the next few months :-)