Automatic Construction Of Finanical Domain Knowledge Graph With Template Induced Method

Knowledge mining is a basic step in the construction of the knowledge map of the financial domain,and the extraction of structured knowledge from the explosive growth of web pages is the main source of data.In extract structured data from the semi-structured section study,based on the template of inducing the way into the mainstream direction,and the current template induction scheme,there is impossible to implement effective shielding noise node information and complex,etc.In this article,template induced technology is used to induce more general templates and use this template to extract structured data.How to integrate data from multiple data sources becomes particularly important after the physical information of the major financial websites and the relationship data between entities.Were introduced in this article,based on the knowledge of the relationship between the entity name and entity attributes fusion scheme,and achieved good results,final figure database,using the secondary will be structured data into database,use of Cypher language query data,making it easy for users to access quickly.Experiments show that the designed template induction program in implementation complexity,node noise resistance,template has obvious advantages,commonality in data fusion scheme,effectively solved the multi-source data redundancy,maximum entity relationship issues such as merger.