Development On Business Intelligence System With The Big Data For Farmers Market

Farmers market left the impression that people can basically be summed up as "dirty", "chaos", "poor". In recent years, the rapid development of the city there are some proposed "upgrading farmers market", " transforming farmers’market to supermarket" and other programs, However, due to high costs and can not popularized in the country, and this shift in approach is only in terms of hardware. In fact, with the development of Internet technology, farmers market should also use the Internet to transform into a new form of existence, to form the "Internet+ farmers’market, we can also use business intelligence technology, internet of things technology to transform the farmers market at the same tiem. For example, by using internet of things technology, farmers’markets can be split up, that every community has a small farmers’market, adopting the community’s electric mode, a third party or the government to unify the management, you can greatly reduce or even avoid people encounter price gouging, shoddy commodities, food safety issues.Price data, consumer data,etc, can uploaded via the Internet to a unified data server,then we can use business intelligence technology to analyze data.This paper has designed a Business Intelligence System prototype with the big data,which includes data source layer, data extract-transform-load (ETL) layer,big data storage layer, the data analysis layer, and the results show layer.Data source layer is divided into two parts, one of the parts is use of o achieve the management of commodity prices by ESLs developed by Foshan City, China and Technology Co., Ltd., the other part is the farmers’market management system, sales data was collected form the management system.Data ETL layer is implemented by the engine of Pentaho Kettle, and has designed an ETL job which can extract data from data source layer.The big data storage layer is implemented by using MySQL Cluster technology, this paper achieved a prototype of a database cluster of four data nodes, two SQL nodes and two Manager nodes.The data analysis layer consists of three parts,the first part through the use of open source relational online analytical processing(ROLAP) engine Mondrian to complete the online analytical processing (OLAP);the second part uses the JaspeRreports report engine to compete the report function;the last part uses the R language to realize the commodity correlation analysis algorithm, but also in the R environment it can complete others needs. The results show layer demonstrates the final results by HTML page, report and chart.The first part,this paper introduces the present situation of the farmers market and current issues facing our farmers market. The second part introduces concepts of big data, business intelligence and electronic shelf label system.The third part introduces the designation of databases prototype and the solution of storage and analysis of big data,the forth part details the big data business intelligence system prototype design, and finally demonstrates some of the features of the system.According to the test, the system can run more stable, capable of handling large-scale data analysis needs,and capable of expanding business intelligence capabilities based this paper.It will bring greater social and economic benefits by applying this system. From social analysis, it protects food security, improves the shopping experience and so on. From the economic analysis, it gets great deal of data, saves operating costs, reduces management difficulties, reduces the attrition.