In view of the construction of the intelligent expert system of the billion point intelligence map, the knowledge map has emerged as a promising tool with high fit. It can lead out, integrate, process and utilize many entities and relationships embedded in products and services, as well as their stakeholders. For this reason, we combine the expert system with the knowledge map to establish an expert system based on the knowledge map. By using new generation technologies such as natural language processing, knowledge graph, topic model, etc., the "problems" and "technical solutions" in intelligent manufacturing knowledge engineering are discovered and standardized. First, we use the top-down construction method, using protégé for ontology modeling, and secondly, through natural language processing and ML to automatically extract entities, relationships and attributes from semi-structured or unstructured knowledge resources, to collect a variety of different forms of raw data for text mining, formal knowledge map.