Based on the semantic analysis technology, knowledge graph is to mine and drill information from input letters, words and articles at multiple levels and dimensions by employing deep neural network, NLP framework semantic understanding and other intelligent processing technologies, using tens of millions of multilingual model libraries as the core and leveraging open data, providing remotely invoked entity extraction, relation extraction, attribute extraction and other algorithm service interface capabilities. The ultimate goal is to build a multi-field knowledge graph platform to serve customers in different industries and application scenarios.