Balanced text recommendation method

The invention discloses a balanced text recommendation method. The method comprises the following steps: acquiring a plurality of balanced training texts; carrying out pre-training on the built Bert-embedding classification model through a training text; classifying the platform text data comprising a plurality of first texts through the trained Bert-embedding classification model; acquiring a first text currently browsed by a user of the platform as a comparison text; extracting a plurality of other first texts in the same category as the comparison text as a comparison text set; calculating a first similarity between all the first texts in the comparison text set and the comparison text; sorting all the first texts in the comparison text set according to the calculated first similarity; and recommending the sorted first texts to the user as recommended texts. According to the balanced text recommendation method, the classification effect of the classification model is better, and meanwhile, related content