Research on Library Big Data Cleaning System based on Big Data Decision Analysis Needs

In the era of big data, university library information management services must be based on actual conditions, using high-quality data to improve big data management. However, high-quality big data is useful data that needs to be filtered and classified. Big data cleansing is an effective way to improve data quality. To this end, the paper proposes to integrate the data resources of efficient libraries, analyze the source and type of useless data, and design a hierarchical management model of data. The model includes management operation level, data cleaning and filtering level, data integration level and big data. At the resource utilization level, after attempting to filter invalid data through data cleaning, the complexity of big data decision analysis is reduced, library big data integration is promoted, big data decision-making is realized, and the possibility of library big data integration and sharing is improved.