[自动翻译] Build and query focused knowledge graph subgraphs locally on resource-limited edge devices for low-latency, offline-aware reasoning and incremental up...