In addition to opening up the business system to obtain data in the main business flow the pattern can also be enlarged. We can also pay attention to data sources other than the companys selfdeveloped systems such as systems purchased from the company common ones such as SAPs industry finance ERP WMS and other systems the data accumulated on the platform when the company conducts online business on thirdparty ecommerce platformschannels the market competition data of the companys competitors the volume of trafficinformation channels related to the companys business.
Public opinion Unstructured data such as user interests and pref Job Seekers Phone Numbers List erences such as Baidu search Douyin etc.. ② Solve the problem of multisource heterogeneity With our efforts data islands have been broken down. With the gradual enrichment of data sources the problem of multisource heterogeneous data has surfaced. It determines the upper limit of data efficiency and the lower limit of data quality. . Anyone who has played the Civilization series of computer games should know what are the landmark events in the era of industrialization and the maturity of industrialization? Parts standardization.
This principle is the same here. The process of solving multisource heterogeneous problems is the process of source data standardization. Solving the problem of multisource heterogeneity in the data collection process is the first level of data standardization work. ③ Source data quality control Speaking of data quality in fact this is a topic in the entire data construction and governance work. You can even build a system specifically to manage data quality which belongs to the category of data management. But why put source data quality control in data collection? Thats because to ensure that the final data quality meets standards the source is the top priority.
發表於 11:48:39
