Building Next‑Generation Data Intelligence Infrastructure with Knowledge Graphs: From New Infrastructure to Cognitive AI Platforms

Introduction – Artificial intelligence is a core driver of the national new‑infrastructure strategy, shifting from perceptual to cognitive intelligence, with knowledge graphs acting as the cornerstone for this transition.

New Infrastructure Overview – New infrastructure includes 5G, data centers, AI, industrial internet, high‑speed rail, ultra‑high voltage, and EV charging; the talk focuses on data centers and AI.

Data Center Insights – China’s data centers account for ~23% of global capacity, with a trend toward larger, more scalable facilities; the sector is poised for rapid growth under policy support.

AI Landscape – Current AI is at the perception level; the roadmap moves toward cognitive intelligence, requiring knowledge integration, reasoning, and decision‑making capabilities.

DIKW Model – The Data‑Information‑Knowledge‑Wisdom hierarchy links data (big data), knowledge (knowledge graphs), and wisdom (AI), illustrating the path from raw data to intelligent applications.

Knowledge Graph for New Infrastructure – Knowledge graphs enable semantic representation, storage, and reasoning, supporting applications such as search, chatbots, recommendation, and smart devices.

Intelligent Data Governance – A unified knowledge representation model (concepts, entities, attributes, relations, events, rules, multimodal links) addresses governance pain points: unstructured data utilization, multimodal fusion, relational exploitation, flexible business models, and intelligent application support.

Knowledge‑Driven Governance Process – Steps include unified modeling, knowledge extraction, multi‑strategy information extraction, deep semantic fusion (ontology alignment, entity alignment, relation discovery, entity linking), and polymorphic storage using graph databases combined with other engines.

Smart Data Governance Platform – The platform leverages the unified knowledge graph to provide semantic search, QA, recommendation, and decision analysis.

Cognitive AI Middle‑Platform – By micro‑service‑izing components, pre‑building models, and enabling business orchestration, the platform achieves rapid, reusable AI solutions across domains such as finance, intelligence analysis, insurance, e‑commerce, and compliance.

Industrial Practice – Case studies include a financial risk‑control middle‑platform, an intelligence analysis middle‑platform, and various domain‑specific applications built via the orchestration engine.

Conclusion – Knowledge graphs bridge data, knowledge, and AI, forming the backbone of next‑generation intelligent infrastructure and enabling fast, scalable, and reusable AI services across industries.