International Business Machines Corporation
Self-learning operational database management

Abstract:

Systems, methods, tools and computer programming products for implementing a cognitive data lake that selects or recommends operational database based on historically created data lakes storing files having similar file types, categorizations, metadata, and/or frequency of file usage. Selecting the appropriate operational databases is streamlined by building a knowledge base that is accessible during the creation and/or registration of each data lake. The knowledge base maintains historical records of past data lakes, describing each historical data lake's file types, attributes, metadata, frequency of file usage and at least one operational database implemented to manage the files of the historical data lake. Incoming files streamed or stored by the data lake are analyzed for metadata (extracted from the files as well as separate metadata), categorized based on the metadata, and matched to one or more operational databases of the historical data lakes described by knowledge base, known to manage similar files.