The transparency of the decision-making process, the credibility of artificial intelligence can be effectively enhanced, and human beings can carry out targeted error correction and further training of artificial intelligence.