In the era of big data, the key to enterprise information resource integration is to rely on enterprise master data management (MDM, Master Data Management), to strengthen the construction of data standardization, and to realize the unity of the information resource model. Enterprise master data management is to integrate the most core and most necessary **** enjoy the data (master data) in multiple business systems of the enterprise, centralize data cleaning and enrichment, and distribute the unified, complete, accurate and authoritative master data to the applications that need to use these data in the enterprise in the form of services. Summarizing years of experience in enterprise informatization planning and combining the requirements of enterprise information resource management in the era of big data, Saidi Jingliu puts forward a methodology to realize enterprise master data management in five stages: identification, diagnosis, planning, implementation and maintenance.

2. Eliminate data heterogeneity

In the strict sense, data heterogeneity is impossible to eliminate, but we can integrate structured data and unstructured data to achieve an elimination through the indirect method. The scientific and effective management of the massive and complex data information of the present enterprise is the premise to guarantee that the big data technology can fully explore the potential value of enterprise information resources. Paper-based information and digital video, audio, mail, pictures and other unstructured data in the enterprise information resources in the proportion of the gradual rise, contains a wealth of potential value. The construction method of these unstructured data has high repetition rate, obvious redundant storage, and there may be complex relationships between different objects. However, the traditional object-oriented data model cannot realize the organization and management of unstructured data. Therefore, enterprises need to promote the integrated development of structured and unstructured data, integrate hypertext, hypermedia data models and object-oriented data models, and construct data models suitable for the unified organization and management of structured and unstructured data.

3. Deploy big data applications

Big data applications, that is to say, data integration systems, data integration software and other applications, enterprises should increase the investment in technology or capital, to set up their own big data team or go to find manufacturers who can provide data integration systems, data integration software services, and even data analysis, data mining and other applications, to build a big data platform.

4. Data security issues can not be ignored

Data integration system, data integration software is the center of the data, and data security issues will affect the entire enterprise, so in the enterprise to use data integration system, data integration software for information resource integration is to ensure that data security as a prerequisite from a variety of aspects, including the management system, processes and so on to ensure that the safety of big data.