Application of intelligent elderly care service in China is still under the trials. Its service effect is poor. Among them, the lack of scientific classification system and public access to smart data resources are one of the main problems affecting the effectiveness of intelligent elderly care services. This paper mainly discusses the originality, composition, characteristics, and classification of data resources in intelligent elderly care service scenarios. This paper states that intelligent care for elderly is a set of activities that continue to occur in a certain period in different scenes. Each type of activity in independent scene of elderly care produces a set of different types of data. This study puts forward the idea of constructing a multi-scene and multi-dimensional intelligent elderly care data resource classification system, based upon use of the faceted classification idea of library science and relevant theories of data management to analyze the source, composition and characteristics of the intelligent elderly care scenario and its data resources, and summarizes the common characteristics of data resources in different scenarios of intelligent elderly care. The study found that the multi-scenario and multi-dimensional intelligent elderly care data resource classification system can facilitate the management and utilization of data resources in intelligent elderly care service scenarios.