Research On Regional Enterprise Data Mart And Risk Identification Algorithm Based On Financial Big Data Platform

Data has become an important asset in big data era.The organic combination of financial data,government shared data and internal system data,the use of big data technology for effective data processing and mining of various types of data,and the timely discovery of the laws and characteristics of financial risks is a topic that needs to be addressed by government departments and financial regulatory units.Based on the financial big data platform construction project in a region,the thesis mainly researches to solve three levels of problems,how to provide a platform basis for regional enterprise risk monitoring research and exploration by establishing a financial big data platform,opening up the financial data of each banking financial institution in addition to introducing regulatory internal data and relevant data from other government departments;how to build an enterprise data marketplace based on the financial big data platform The paper also discusses how to provide a data base for enterprise risk monitoring data analysis;and how to solve the difficult problems of enterprise risk monitoring by mining enterprise relationship data.In response to the above three problems,the main work done in the thesis is as follows.Design a financial big data platform that meets the needs of regional financial economic development and enterprise risk monitoring,focusing on the data mart support architecture under the financial big data framework.Propose a model for building enterprise relationship data mart based on knowledge graph,which mainly solves the defects of inflexible modeling and slow multi-dimensional query of traditional data mart modeling in enterprise relationship data mart,and builds enterprise relationship data mart based on data collected by regional financial big data platform through the process of data selection,entity extraction,attribute extraction and relationship extraction.We analyze the current problems and difficulties of guarantee circle identification and key enterprise monitoring in enterprise risk monitoring,design the enterprise relationship network construction model based on enterprise relationship data mart,and propose the AHPL guarantee circle identification algorithm based on enterprise relationship network,which mainly solves the problems of single data and loose network in guarantee circle identification.The test results show that the algorithm can combine guarantee circle identification with relationship data other than non-guarantee relationship data,and can better identify hidden guarantee circle;through the improvement of PageRank algorithm,the key enterprise monitoring identification algorithm based on the improved PageRank algorithm is proposed,and the test results show that the result data of the improved algorithm presents more loose in indicators,and can better find the key monitoring enterprises with higher risk.