Intelligent Information Analysis Framework TjuIAF And Empirical Analysis

A new era of intelligence with the interconnection of human and physical objects into a new production relationship face many new challenges of information analysis,such as deepfake,fake news and so on.It is necessary to reconstruct the information analysis framework based on big data and AI,which is to integrate many kinds of data of social space,physical space,information space,and cognitive space.The complex network algorithms will be used to perceive,understand and predict the targets including people(fake),things,organizations,true and false events and their interrelationships.This paper focuses on the research of information analysis framework tjuIAF and achieves the following results:First,a framework for information analysis(tjuIAF)is constructed.According to the specific information needs,the data from social space,physical space,cyberspace and psychological space is collected and cleaned.Then,multilayer heterogeneous complex network is constructed built on the processed data to complete information fusion.Next,based on the results of information fusion,the ways of information analysis including information perception,information understanding and information prediction are carried out to provide guidance and suggestions for practical needs.Secondly,the paper uses the data of a commercial bank to demonstrate the key nodes identification method in tjuIAF framework for the identification of important customers.This method combines operational attributes with network topology attributes and is a judgment index to evaluate the importance of customers.Topological structure attributes can describe the key degree,mediation degree and connection degree of nodes in the network.In addition,operational attributes can personalize the importance of objects in a specific industry.The combination of the two attributes can describe the research object,that is,the importance of the customer to the industry.Finally,This paper uses the data of a commercial bank to demonstrate the risk role identification method in tjuIAF framework for the early warning of customer loss.For a commercial bank,the time section is delimited to calculate the trend of customer importance over time to determine whether the customer has the risk of loss.The results obtained by this method have been proved to be consistent with the actual scenario of a commercial bank,which can help it to locate customers at risk of loss and take measures in advance.Applying this intelligence analysis framework to the field of commercial Banks,information analysis of commercial Banks' customers can be realized,and risk management and precision marketing of commercial Banks' customers can be assisted.Not only for commercial Banks,the framework has good scalability,suitable for the application industry that needs accurate perception of customer intelligence,and has certain realization value and application prospect.