Credits: 6 Cours: 2h/semaines Projets: 2h/semaines Prof. Pierre Morizet-Mahoudeaux et Jean-Benoist Leger 2019

This course aims at presenting the principles of tools development and using for data warehouse conception and decision taking with specific tools (Business Objects, regression, segmentation). The class NF26 aims at presenting the data warehouse construction principles, with an introduction to NoSQL. In this class, a first part is focused on the major concepts related to data warehouses. The students learn how to analyze “clients” needs and requests so as to conceptualize an architecture that will answer those needs in an optimal, simple and ready to use way: the data marts. In this part, the differences, pros and cons of different data warehouse models are emphasized. In order to conceptualize a data warehouse, the students learn to separate business process data into facts, holding measurable and quantitative data about a business, and dimensions which are descriptive attributes related to fact data. Examples of fact data include sales price, sales quantity, and time, distance, speed, and weight measurements. Related dimension attribute examples include product models, product colors, product sizes, geographic locations, and salesperson names. Main theory of the first part includes the following keywords: dimensions tables - facts tables - hierarchy - normalization - denormalization - star schema - snowflake schema.

In a second part of this class, students learn a new paradigm for data warehouse implementation and conceptualisation : NoSQL. Using Cassandra technologies and Python as main programming language, students implement a complete data warehouse following NoSQL and column table logic. In this project the students will take advantage of Cassandra model for presenting a dataset in a way that is interesting for analytics, data visualisation and that allows non-technical people to retrieve and get real business value on top of the row data set. Main theory of the second part groups the following keywords: Cassandra - CAP Theorem - Eventual consistency - NoSQL - Consistent Hashing - Nodes - ACID and BASE - Analytics - Kmeans.

Technologie : Pentaho, Birt

Technologie : Python, Cassandra, Spark