Design a Data Warehouse Schema from Document-Oriented database

Abstract

Traditional data warehouses are unable to meet the growing needs of the modern enterprise to integrate and analyze a wide variety of data generated by social, mobile and sensor sources. What is remarkable is that many companies have changed their data storage in NoSQL databases. Given the important role played by this type of databases, it is necessary to study NoSQL databases as a data source for the modeling and the implementation of data warehouses. This paper proposes a method to design the data warehouse schema from schema free databases known as NoSQL databases. Our proposal starts with the extraction of schemes from document-oriented databases as an example of NoSQL database. This extraction is performed based on the MapReduce paradigm. Then, it defines the structure identification graph from the extracted schemes. This graph will help the designer to identify the multidimensional concepts for each schema extracted in order to design the global data warehouse schema.