基于知识图谱的金融信息系统动态管理数据更新查询仿真

Simulation of Dynamic Management Data Update and Query in Financial Information System Based on Knowledge Map

金融信息系统内存在多种类型的通用管理数据,这些数据在实际仿真更新查询过程中会产生干扰,输出的冗余数据会导致仿真数据的丢失量较大。针对该问题,文章设计了基于知识图谱的金融信息系统动态管理数据更新查询仿真过程。其采用模式图储存类型定义金融信息系统内的抽象指令关系,定义金融信息系统知识图谱邻域,采用LSTM模型内的遗忘门丢弃知识图谱邻域内无用的信息,排除冗余数据的干扰,输出系统动态管理数据;采用向量空间模型描述管理数据以及其对应的管理对象,根据查询文档与更新向量之间的嵌入关系,仿真得到更新查询过程;爬取采集金融信息系统管理数据,编码并清洗动态管理数据。测试结果表明,通过该方法可有效采集管理数据,实际应用性较好;采集到的管理数据多数可更新为互信息的形式,仿真丢失量为5M,数据丢失量较小;仿真管理数据更新查询过程更加真实,仿真更新查询过程真实性较强。

There are various types of general management data in financial information systems,which may cause interference during the actual simulation update query process,and the output of redundant data may lead to significant loss of simulation data.To solve this problem,this paper designs a simulation process for dynamic management data update and query in financial information systems based on knowledge maps.It uses a pattern graph storage type to define abstract instruction relationships within the financial information system,define the knowledge map neighborhood of the financial information system,use a forgetting gate within the LSTM model to discard useless information within the knowledge map neighborhood,eliminate the interference of redundant data,output system dynamic management data,use a vector space model to describe management data and its corresponding management objects,and update the embedded relationship between the query document and the vector,Simulate the update query process.Crawl and collect financial information system management data,code and clean dynamic management data.The test results show that the article method is effective in collecting management data and has good practical applicability.Most of the collected management data can be updated in the form of mutual information,with a simulation loss amount of 5M and a minimum data loss amount.The simulation management data update query process is more realistic,and the simulation update query process is more authentic.