Quality human resources to support productivity performance is very important, the Indonesian government through KOMINFO which is in charge of the communication and informatics sector which focuses on the development of the telecommunications sector, internet governance and digitalization has a program which one of its programs is to improve human resources through the Digitalent program. The Digitalent Program has been a competence development training program since 2018 with the main objective of increasing skills and competitiveness, productivity, professionalism of human resources in the field of information and communication technology for Indonesia's young workforce, the general public and the state apparatus, for this reason it is necessary to conduct research to find out how is the public sentiment towards the Digitalent program, does the majority of the public assess the Digitalent program with positive sentiments or judge it with negative sentiments. The large amount of review data requires a long time to read one by one, therefore public response regarding the Digitalent program was obtained by scraping techniques on Twitter from January 1 2018 to September 1 2022 by producing 9934 data. This research used preprocessing cleaning text, tokenizing text, normalization, filtering, to sentence, stemming text. The Naïve Bayes classification used is the Naïve Bayes Multinomial text classification and the results of the classification accuracy using the Naïve Bayes Classifier Multinomial produce an accuracy of 79% with the highest precision of 84%, the highest recall of 89% and the highest f1-score of 86%.