Using 1 531 comment data of Sina Weibo and 2 587 of Baidu index data from 2013 to 2020, based on word cloud analysis, text sentiment analysis, and ARIMA-SVM prediction model, based on word cloud analysis, text emotion analysis and ARIMA-SVM prediction model. It was found that residents most concerned about the nutrient reference value of energy, and the nutrient reference value of moon cakes, soy sauce, yuanxiao, milk and other products were more used, especially children's food. In general, the nutrient reference values obtained slightly higher praise, but there were also many controversies; nutrient reference values did not get good attention in the network, and did not increase with time; from 2020 to 2022, the enthusiasm of the nutrient reference values to be concerned will still be not high. Therefore, it is urgent to improve the popularization of science, add the nutrients that are concerned, and implement the FOP labeling combined with the nutrient reference values to promote to nutrient reference values play its due role.