User dynamic interest recommendation method and system based on recurrent neural network

The invention relates to a user dynamic interest recommendation method and system based on a recurrent neural network, and belongs to the technical field of recommendation. The system comprises a data preprocessing module, an embedding module, a periodic interaction module, a multi-interest extraction module and an interest evolution module. The data preprocessing module obtains historical interaction item data of a user to form a historical interaction item sequence of the user; the embedding module converts a high-dimensional sparse feature vector into a low-dimensional dense feature vector; the periodic interaction module brings periodic information into interest representation of a user, and designs a graphic structure to capture global and local interactivity between projects; the multi-interest extraction module extracts multiple interests from the user sequence based on a self-attention method; the interest evolution module uses the AUGRU to capture an interest evolution process associated with the tar