Why NeuralKG

The knowledge graph stores a large amount of knowledge represented by symbolic triples. With the maturity of automatic extraction technology and a lot of human input, many high-quality knowledge graphs have been successfully constructed, and are used for search recommendation, intelligent question answering, root cause tracing, etc. Application tasks provide rich symbolic knowledge services. With the rise of deep learning, the algorithms of many downstream tasks are calculated based on vector space, and the symbolic knowledge service also brings additional tasks of symbolic knowledge representation and fusion to downstream tasks. Therefore, we have developed the NeuralKG tool to integrate mainstream The knowledge graph representation learning framework is included, which is a general tool for the vectorized representation of knowledge graphs.