Font Size: a A A

Design And Implementation Of Intelligent Financial Wallet System Based On Microservice

Posted on:2021-05-01Degree:MasterType:Thesis
Country:ChinaCandidate:B WuFull Text:PDF
GTID:2518306503999479Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Financial wallet means money which customers deposit in a mobile wallet system.It's an Internet-based financial system which the customers use for mobile payment,settlement of bills,money or fund management,consumer credit service and other financial value-added services.For example,Alipay's Yu'E Bao is the most representative one.Intelligent financial wallet is another money management wallet system which uses machine learning technology to increase the yields from money funds.As result of the increasing popularity of Alipay's Yu'E Bao,there is a booming development of numerous money management wallet systems like Yu'E Bao in the market.At present the market volume is as high as RMB 7 trillion.However,with the rapid expansion of the money fund market,the regulatory authority has issued a series of regulatory policies and ordered relevant enterprises to make rectification within a time limit.The adoption of microservice architecture can support the rapid reconfiguration of the system and the adjustment of business processes under the complex and changeable market environment and policy factors.However,there is no general specification and method of microservice boundary division in the industry,and if the microservice boundary division is unreasonable,the following problems will be caused:(1)interface coupling between the service end and the consumer end;(2)too many coarse functions of the service grain make it difficult to maintain and expand;(3)too fine service grain makes the operation and maintenance complexity increase;(4)cap related problems such as distributed transactions may be introduced Question.On the other hand,the yields from money funds are continuing to decline.For example,their 7-day annualized returns generally are falling from over 6% in 2014 to less than 3% at present.And because the regulatory authority limits the liquidity of money funds after imposing a limit on daily T+0 withdrawal from the money funds,the attraction of money management products is reduced considerably.How to improve the liquidity of their products and overall returns in the current policy environment is an issue that many Internet-based financial companies need to consider.Regarding the question and issue mentioned above,this paper,which takes an Internetbased financial company's project as the background,has designed and built an intelligent financial wallet system(hereinafter referred to as IFW system)based on the micro service architecture after making a deep research on micro service,workflow,machine learning and other related technologies.After being tested and putting into application,the results show that the system is feasible and effective.The main points of the paper are as follows:(1)As the business process of financial wallet needs to be rapidly adjusted and restructured according to the market and policy environment,the paper makes an analysis of the technologies,such as the means of demarcating the boundary of the micro service architecture,the development framework,the communication and invocation mechanism,the deployment mode and workflow technology.After summarizing the features of microservice,the paper puts forward a means of demarcating the boundary of the micro service architecture based on business entity(DBME).It also puts forward a mechanism of asynchronously choreographing microservice based on a workflow engine,and uses the Zeebe workflow engine to implement the mechanism.During the half year trial operation,results show that the system is able to complete the rectification and testing within two weeks or one month before going on-line according to the amendment request made by the customer.(2)As for the issue that the yields from money funds are trending down as a whole,this paper applies the quantitative investment idea to investment in money funds.After analyzing the mainstream quantitative investment models and machine learning algorithm,a penalty linear regression algorithm is proposed to predict the return of multiple monetary funds.By continuously adjusting and optimizing the proportion of different funds,the fund with low return and the fund with high return can be sold to improve the overall return on investment.the results show that the yields from money management products increase by between 15%and 20% compared with the investment in Yu'E Bao.(3)As for the issue that central bank reduces the liquidity of money fund products after imposing a limit on daily T+0 withdrawal from money fund product,zhe paper,after analyzing the customers' monetary demand and behavior,build a user profile of a merchant and using ARIMA autoregressive moving average algorithm to predict the money demand of businesses.this paper puts forword an intelligent investment model of money fund(IIMMF).The model takes the predicted fund return and merchant's capital demand as the eigenvalue to construct the multi-fund portfolio.the results show that the upper limit of daily T+0 withdrawal from money management products based on this model increases by more than 500%.
Keywords/Search Tags:Intelligent financial wallet, Microservice architecture, Workflow engine, Machine learning, Quantitative investment
PDF Full Text Request
Related items