[Purpose/significance] Integrating information field theory from the perspective of user information behavior, it is to analyze the altmetrics data generation mechanism, aiming to clarify the core logic of data generation and promote it as a reliable bridge connecting science and society. [Method/process] Facing the scientific information communication process based on the network platform, the information field theory is introduced, and the stakeholder needs involved in the communication process are taken as the basis to construct a "Context-Cognition-Behavior-Data" (CCBD) analysis framework, which take the platform as the information field and interpret the generation mechanism of altmetrics data. [Result/conclusion] The network information platform is a virtual information field for scientific information communication. The altmetrics data generation process from the perspective of user information behavior includes five stages: triggering, acquisition, judgment, utilization, and dissemination. Four types of generation mechanisms of altmetrics data under the CCBD framework are revealed: stimulus-disturbance, competition-selection, overlap-self-organi-zation, and consensus-collaboration. The four types of mechanisms do not exist independently, and present a hierarchical progressive relationship, which are "micro triggering-meso selecting-macro aggregation-multiparty collaboration".