Lecture Slides

Slides are subject to change until shortly before the lecture when they're delivered!
  1. Introductory slides (Sept 8) also with pauses
  2. Data Warehousing slides (Oct. 22) also with pauses
  3. DW Issues (by examples) slides (Oct. 27) also with pauses. Daniel made some lecture notes too.
  4. Views and their maintenance in DWs (Oct. 29) also with pauses
  5. Materialized view selection for DW (Nov. 3 & 5) also with pauses
  6. OLAP motivation; storage schemes; snowflake (Nov 10) also with pauses
  7. OLAP operations (Nov. 12) also with pauses
  8. Multidimensional query languages (via MDX) (Nov. 17 &19) also with pauses
  9. XMLA and JOLAP (Nov 19&24) also with pauses
  10. Data Mining (Nov 24 or 26) also with pauses
  11. Range Queries and Wavelets (time permitting) also with pauses

Homeworks

  1. Assignment 1A, due 10 November.
  2. Assignment 1B, due 19 November.
  3. Assignment 2 (final version), due 1 December, with provisions for late submission (with penalty) until 10 December.

What do the UNB administrators use?

It appears the use OLAP, via a graphical ASG-Safari.OLAP product.

Link to Mondrian (an XMLA provider)

Main Mondrian page on Ennui For the XMLA tester, you need this line to replace the DataSourceInfo line in the queries.
version of FoodMart.xml that corresponds to Mondrian on ennui. This gives the schema of the FoodMart cubes (before a new cube was added for Assignment 2).

Test information

The final test will be held 3:30 - 5pm on December 1 (the location was announced in class). It covers the second half of CS6905.

The first question presents a short data warehousing/OLAP scenario and asks 8 questions, mostly about how you would do various things, what some difficulties might be, and so forth. This question is worth 55%. Most of the 8 questions can be answered in one or two sentences; if you write big paragraphs, you will run out of time. Many of the questions would have many different right answers (and even more wrong answers...)

Then there are 4 more questions, worth 45% in total. One question requires that you write some SQL. However, there are no detailed MDX questions on the test. I do not believe that you would need to memorize any complicated formulae, algorithm or terminology. (However, this doesn't mean the test is easy.)

You don't need to read most of the slides on Range Queries and Wavelets, but you should have some idea of the basics of the "prefix sum" method explained on the board. Also, you don't need to know about how to "compute the data cube" (slides at the end of the "OLAP Operations" bundle). Other than that, the content will involve things discussed in class or in the homeworks.

Papers Referenced

  1. The paper by Johnson should be on reserve in the Engg library.
  2. Most other references can be found in the Data Warehousing and OLAP Bibliography which Daniel has been maintaining

Valid HTML 4.01!