The recommendation of the Next Point of Interest (POI) for human trajectory prediction is crucial across various applications, including travel recommendations, location-based advertising, and traffic management. However, accurately forecasting behaviors in real-world scenarios poses significant challenges due to: 1) the high-order and scattered nature of movement data encompassing both temporal and spatial contexts, which constrains prediction accuracy; and 2) the complexity of behavior prediction tasks, influenced by factors such as historical data, regional environment, and population effects. This paper presents a hypergraph-based neural network framework called Hyper-Intent POI Recommender (HIP) to tackle these challenges. In accordance with the Movement Ecology Paradigm, we encode three internal and external factors including regional activities, personal historical records, and group movement patterns by adopting a hypergraph approach. Subsequently, we utilize a graph attention neural network to encode points of interest (POI) nodes and integrate these factor embeddings for enhanced location prediction. Our method improves the quality of embeddings for influencing factors, and comprehensive experiments conducted on three publicly available datasets demonstrate that our model surpasses other baseline approaches (e.g., sequential-based and graph-based models) in terms of top-K accuracy and mean reciprocal rank (MRR) for Next POI recommendation tasks.