Abstract: Objectives:  Personalized point-of-interest (POI) recommendation is a vital service in location-based social networks (LBSNs). It can effectively use the sequence and spatiotemporal context information of check-in data to discover users' movement patterns and preferences.   Methods:  This paper proposes a probabilistic generative model with embedded spatiotemporal conditions to fully exploit the long-term dependency between users' personalized spatiotemporal preferences and sequential check-in sequences, constructs a gated dilation residual network, and implements a POI recommendation method based on gated dilation residual network. The method in this paper learns users' check-in sequences through a gated dilation residual network. It mines and captures the spatiotemporal patterns, sequence preferences and temporal preferences of users' sequential check-in behaviour using the spatial distance and time interval of users' sequential checkins as constraints.   Results:  The proposed method in this paper shows significant improvements on the Foursquare and Instagram datasets. Compared to the best-performing algorithm NextItNet, our method demonstrates noticeable enhancements in terms of recall, precision, F1 score, and normalized discounted cumulative gain (NDCG). On the Foursquare dataset, we achieve improvements ranging from 1.52% to 24.95%. On the Instagram dataset, the improvements range from 7.06% to 42.47%.   Conclusions:  The proposed method is more suitable for mining the longterm dependency relationships in continuous user check-ins. It effectively incorporates spatial distance and temporal interval factors, thereby improving the accuracy of POI recommendation.