Assessing and Refining ChatGPT's Performance in Identifying Targeting and Inappropriate Language: A Comparative Study