Telecom operators have accumulated data for many years, with structured data such as financial income and business development, as well as unstructured data such as pictures, text, audio and video. From the perspective of data sources, the data of telecom operators comes from all businesses involving mobile voice, fixed telephone, fixed network access and wireless Internet access, and also involves public customers, government and enterprise customers and family customers. Meanwhile, they also collect contact information from all types of channels, such as physical channels, electronic channels and direct channels.
On the one hand, massive data brings a large number of information assets to operators; on the other hand, useless data and interference data also show an exponential increase. The quality of data is directly related to the accuracy of information. If the problem of data quality is not solved effectively, data assets will not be able to effectively reflect the facts of enterprise operation and market, and business decisions will lose reliable basis. Therefore, centering on the life cycle of enterprise big data, it is particularly important to realize the whole process quality monitoring of key data.
Through the construction of data quality evaluation system, the implementation of quality inspection, analysis and monitoring, and the establishment of a continuous improvement mechanism, the big data solution of Whales Technology can realize the comprehensiveness, controllability and measurability of data quality control, quickly locate and effectively solve quality problems, and proactively discover hidden risks of data quality.
I. Standardized quality audit model
Built-in dozens of quality assessment technologies, including data output time timeliness, output interval timeliness, primary key repetition, non-empty decision, value range, total consistency, summation consistency, field detail consistency, to meet the definition of various business data detection rules; Flexible extension through plug-in mode, to adapt to the current and future data quality management requirements of enterprises.
Second, closed-loop efficient problem handling mechanism
Dynamically balance orders and alarms after problems occur. The whole process of problem processing is tracked visually and transparently by the system. The experience of problem handling is summarized, refined and summarized to form quality knowledge and precipitate to the knowledge base for sharing.
Three, multidimensional comprehensive analysis
Through data statistical analysis, visual data perspective analysis and other technologies, automatically generate quality analysis views, data quality reports and data problem handling efficiency, assist users in quality analysis of problem data, and provide strong support for users to carry out targeted quality improvement.
In 2018, the data quality management system was launched in Guangxi Telecom, which automatically found and processed more than 99% of data quality problems every day, effectively improving the reliability and reliability of data; Due to the high degree of automation, the product reduces the manpower input of operation and maintenance, greatly shortens the management cycle, effectively reduces the data management cost by 60%, and escorts the clean and high-quality data assets.
