Classification method based on big data

The invention discloses a classification method based on big data. The classification method based on the big data includes the following steps: establishing a trained model; acquiring news D, and carrying out work segmentation on the news D to obtain news S; and obtaining a classification category of the news S through the trained model. The specific training steps of the trained model include: carrying out work segmentation on news D1 to obtain news S1; establishing a VSM space vector model for the news S1, and extracting feature words t; calculating correlation probability of the news S1and all categories on the basis of the feature words t; judging whether calculation is completed, if yes, obtaining a category c with maximum correlation probability, and otherwise, continuingcalculation until calculation is completed; classifying the news S1 to the category c; and judging whether training is completed, if yes, saving the trained model, and ending training, and otherwise, continuing next training. 本发明公开了种基于