Next interest point recommendation method based on adaptive space-time diagram representation learning

The embodiment of the invention provides a next interest point recommendation method based on self-adaptive space-time diagram representation learning. The method comprises the steps that user identity information and user access place information are collected; taking the user identity information and the user access place information as input, and obtaining geographic map semantic information based on a preset embedded gate controlled geographic map surveying and mapping model; taking the user identity information and the user access place information as input, and obtaining sequence diagram semantic information based on a preset sequence diagram surveying and mapping model controlled by an embedded gate; and based on semantic joint framework learning, according to the user access place information, the geographic map semantic information and the sequence map semantic information, realizing consistent distribution of the geographic map and the sequence map. The technical scheme provided by the invention is