Application Of Data Lake Based On Cloud In Dealer Index System Of An Automobile Enterprise

With the development and popularization of computer technology,the era of big data has also come.With the continuous development of data technology,all walks of life pay more and more attention to data.In addition to the growth of unstructured data,the amount of unstructured data needs to be increased from the original production level to the original production level.At the same time,with the emergence of cloud computing,cloud storage technology has become an important way for enterprises to store massive data.China's traditional manufacturing industry involves a wide range of businesses and accounts for a high proportion of GDP.However,the process of enterprise digital transformation by using electronic information technology such as cloud computing in traditional manufacturing industry is slightly slow.Especially in the traditional automobile industry,its business scope includes automobile production,parts procurement,automobile sales and automobile aftersales maintenance.These fields are not only closely related to everyone's life,but also have made great contributions to the development of national GDP.The whole business process of the automobile industry will produce a huge amount of data,but there is no unified way for the whole industry to store and use these data efficiently.Taking D automobile enterprise as an example,this paper makes an in-depth analysis of its business process: in the actual operation,the business side of D automobile enterprise needs a large amount of data every day as the support of management decision-making.Although D automobile enterprise has a certain foundation in data automation,there is no good way to save and manage the homologous and heterogeneous data of vehicle sales and after-sales.Many data requirements on the business side are realized by manual calculation.The calculation logic of many indicators is the same every month or quarter,but the time range is different,and the extraction of multi-source heterogeneous data also increases the difficulty of work.Therefore,the work content of the business side in calculating data indicators is cumbersome and repetitive,and the work efficiency is low.How to make more efficient and convenient use of data to provide support to the business side is the main problem faced by D automobile enterprise.In view of the above investigation and research,this paper will build an enterprise level data platform for D automobile enterprise to summarize the homologous and heterogeneous data of D automobile enterprise,and calculate the data indicators required by the business side according to these summarized data.The construction of this data platform is based on the data Lake theory,and the platform is developed by the combination of data warehouse and data analysis technology.The whole development framework is based on Microsoft's azure,data storage is based on HD insight's hive data warehouse,and code development is carried out on databricks.The specific process is as follows: firstly,sort out the business process of D automobile enterprise,and sort out relevant business indicators,namely KPI,in combination with the needs of specific business side.Then,according to the sorted KPI logic,the data lake is operated hierarchically through ETL Technology.The data flows in different layers of the data lake to complete the whole life cycle,and finally get the KPI required by the business department.Finally,combined with power Bi technology,all KPI values are visually displayed.Finally,the data Kanban containing different business modules and different KPIs is obtained,and the automatic refresh function of KPI values is realized.Provide support for business side decision-making.