Databases course by Timothy Griffin

Dr Timothy Griffin had earned his BS in Mathematics from the University of Wisconsin, Madison. PhD in Computer Science from Cornell University. Previous experience includes teaching at UNICAMP in Brazil and more than a dozen years at Bell Laboratories and AT&T Research. Joined the Computer Lab on January 1, 2005. Dr Griffin is now a lecturer at the university of Campbridge, UK.

The aim of the course “Databases” is to cover the principals of databases as seen from the perspective of application writers. The course covers schema design techniques, SQL, data warehouses, On-line Analytical Processing (OLAP), federated databases, and the “NoSQL” movement.

Course Materials:

1. Introduction. What is a database system? Database systems are more than just a collection of data. Three level architecture. OnLine Transaction Processing (OLTP) versus OnLine Analytic Processing (OLAP).
2. The relational data model. Relations are sets of records. Representing entities and relationships as relations. Queries as derived relations. Relations are the basis of SQL (but note the use of multi-sets).
3. Entity-Relationship (E/R) modelling. A bit of set theory. Entities have attributes. Relations have arity. Database design and data modelling.
4. Relational algebra. Relational algebra as an abstract query language. Core operations – selection, projection, product, renaming, and joins.
5. relational calculus Relational calculus as an abstract query language that uses notation from set theory. Equivalence with relational algebra.
6. Schema refinement I. The evils of redundancy. The benefits of redundancy. Functional dependencies (FDs) as a formal means of investigating redundancy. Relational decomposition. Armstrong’s axioms and Heath’s Rule.
7. Schema refinement II and Normal Forms. Schema normalisation. First and Second normal form. Third normal form and Boyce-Codd normal form. Multi-valued dependencies (MVDs) and lossless-join decomposition. Fourth normal form.
8. Transaction management
9. On-line Analytical Processing (OLAP). When to forget about data normalisation. Beware of buzz-words and the Data Warehouse Death March. More on OLTP versus OLAP. What is a data cube? Data modelling for data warehouses: star schema.
10. More OLAP – Dimensional modeling
11. SQL and integrity constraints. An overview of the core of SQL. SQL has constructs taken from both the relational algebra and the relational calculus. Integrity constraints as special queries, often required to yield a null result.
12. Further relational algebra, further SQL