Research And Application Of Subgraph Matching Query Method For Domain Knowledge Base

With the continuous promotion of information construction in various fields,the field data is increasing day by day.Effective organization,mining and utilization of industry data have become the core elements to enhance the competitiveness of various fields.In order to solve the problem of the validity of vertical domain knowledge acquisition,an approximate subgraph matching query method based on knowledge base is proposed in this paper.By using the structural and semantic information of the nodes in the knowledge graph,the filter rules of the nodes are defined,and the filter of the noise data is realized.The matching cost of query node and candidate node is calculated,and the matching cost of node weight to neighbor vector is set to reflect the structural characteristics of query graph node.The sorting mechanism is invoked to further process the query results and return the first k query results that satisfy the user’s query intention.Based on the proposed subgraph matching query method,an experimental platform of agricultural knowledge query system was constructed.The main research contents include:(1)Designed a quantitative model of agricultural knowledge.The core idea of knowledge vectorization is to consider the relationship between entities in the knowledge base while learning text corpus in FastText model.Knowledge vectorization is used to lay a foundation for semantic measurement of subgraph matching and improve the accuracy of subgraph matching.(2)Proposed a subgraph matching algorithm based on node filtering.The algorithm takes into account both structural information and semantic information of nodes to filter nodes,then sets node weights to calculate node matching costs,and invokes sorting mechanism to input and return the first K results.(3)Design and implement agricultural knowledge query system.On the basis of algorithm research,this paper designs and implements an agricultural knowledge query system.The system USES Django as the development tool,MySQL as the database,and realizes the agricultural knowledge query,related search two functional modules.The system converts the related algorithms into visual operation interface and verifies the validity of the related algorithms.