Interest activity recommendation method based on geographic grid

An interest activity recommendation method based on a geographic grid comprises the following steps: dividing a research area into regular grids, and establishing a personal interest grid area for each user through sign-in frequency and preference deviation ratio parameters so as to deduce space activity preferences of the users; capturing other similar user activity preferences by adopting a non-negative tensor decomposition method, and cooperatively establishing time activity preferences of the users; and fusing the space activity preference and the time activity preference by adopting a context-aware fusion method to jointly determine the interest activity recommended to the user. According to the model, the sparsity of user sign-in data is improved based on a geographic grid and a tensor decomposition method, quantitative analysis is carried out on user interest activities, the interest activity recommendation accuracy is improved, and the recommendation result meets the personalized requirements of users