At the beginning of 2020, the sudden outbreak of Covid‐19 quickly caught people's attention. With the increasing number of cases and endless information coming out from the Internet, the negative emotions of netizens are gradually increasing. But at the same time, thanks to the government's quick response and the positive guidance of the media, the public's negative sentiment has been largely eased. Therefore, this paper tends to observe the change of the sentimental color of words on the netizens' attitude towards the COVID‐19 from the government's response and media reports. By searching keywords in the corpus, this work analyzed and compared the sentimental color change trend of words from 2015 to 2019 and 2020. At the same time, this study also uses questionnaires to assist the conclusions drawn from the corpus. The study found that the conclusion drawn from the corpus was that the public's attitude towards COVID‐19 was generally negative, but the questionnaire showed more positive aspects. At present, researches on the COVID‐19 at home and abroad are rarely based on corpus, which is a research blank. Therefore, this paper will study the COVID‐19 from the perspective of corpus, which is an innovation point.