Graph database query statement construction method based on natural language semantic transformation

The invention discloses a graph database query statement construction method based on natural language semantic transformation, which comprises the following steps of: 1) classifying graph database nodes according to node tags, performing entity extraction on a classification result, and establishing an abstract entity set E; 2) utilizing synonym mining to convert abstract entities into corpus entities, and establishing a corpus entity set C; 3) defining a basic graph database query statement template; natural semantics are selected with the help of the knowledge base, and the natural semantics are converted into formal semantics; 4) traversing the semantic entity set Q in sequence, and constructing a query clause pql; the graph database retrieval service construction method based on natural language semantic transformation can be suitable for various different types of graph databases and is good in universality. 本发明公开了一种基于自然语言语义转化的图数据库查询语句构建方法，包括以下步骤：1)按节点标签对图数据库节点进行分类，对分类结果进行实体抽取，建立抽象实体集E；2)利用同义词挖掘将抽象实体具化为