Method for constructing user space-time portrait array based on big data algorithm

The invention discloses a method for constructing a user space-time portrait array based on a big data algorithm. The method comprises the following steps: collecting all-day operation behavior data of different platforms and different users in different systems; preprocessing the collected data to extract user behavior characteristics; establishing user behavior preference labels according to the user behavior characteristics, and performing multi-level division processing; defining user behavior preferences of different time dimensions in a fixed period by using a label attenuation algorithm; analyzing the operation behaviors of multiple users at different time points in batches, and extracting the same service operation behaviors of different users in the same time window; and constructing a service-related user space-time portrait array. According to the method, the time factor and the space factor of the user behavior are comprehensively considered, a dynamic and high-dimensional user space-time portrait