The invention discloses a content-based user personalized product matching recommendation method. The problem that in the prior art, a recommendation algorithm still needs to be improved is solved. The method comprises the following steps: step 1, a random batch sampling method based on a user; step 2, a content-based user product matching method: establishing a network; according to a batch inputuser-based random batch sampling method, obtaining an ordered set of a user id, a historical record product id list, a target product id and label set, and training, adjusting and evaluating a network model on the training set, the verification set and the test set respectively; and inputting a specific user and historical records thereof, predicting scores of the specific user for all unwatchedmovies by using a content-based user product matching network, sorting the scores, and finally outputting a top-N recommendation result. According to the method, the lightweight neural network is adopted, so that the training time and training equipment requirements are greatly reduced, the sampling process is easy, the model input is more random, and the generalization ability of the prediction result is stronger.