Data Warehousing and Business Intelligence Project
Task: Architect, Populate and Explore a Data Warehouse for Business Intelligence
This CA will assess your ability to apply the theories, methodologies and strategies tackled in the course to successfully implement a data warehouse to support business intelligence queries.
CA Deliverables
For the project, there are 2 deliverables: 1) a 20 page report, and 2) a pre-recorded video demonstration with voice commentary of up to 5 mins.
The paper should be structured as follows:
1. Introduction: description of the setting of your data warehouse. This section should document: the sources of data, motivation for these sources of data, and describe how the data sets have been curated as well as coerced to work together. If you have fabricated your own data for this project, describe how this has been done, and what assumptions and/or data artefacts are interwoven in this process.
2. Architecture and Implementation: a high-level overview of how the data warehouse is built, i.e. have you followed in an Inmon, Kimball, or alternate approach. Detail why you have designed the data warehouse in this manner. Use supporting argumentation and (where appropriate) substantiating literature with respect to the architectural, methodological and design choices made. Include any details of staging in this section.
3. Data Warehouse Data Model: description of the schema(s) built. What dimensions do you have, and why? Also include a discussion on why they are composed in this manner. Noteworthy, also is how the data model has been prepared for the BI queries.
4. ETL strategy: how has the data been extracted, transformed, cleaned, and loaded into the data warehouse? Document precisely how the data goes from the sources to data warehouse, ideally as a logical data map. Document all transformations and steps taken to clean the data sources. Note which parts of your approach are automated.
5. Application of Data Warehouse: Execute three non-trivial queries that highlight the business value of your data warehouse. For each query, note the sources of data contributing to the answer of the query. Each BI query should be articulated as a question that is answered using the data warehouse. You may use visualisation tools to execute the queries, e.g. Power Pivot, Tableau, Power BI etc. However, the video MUST illustrate how these tools have been integrated into your overall solution - at least how the cube has been imported/exported.
The final report should not be more than 20 pages including all figures and references.
The video should be no more than 5 minutes in length. You may use any screen capture software,
Snagit, Camtasia, CamStudio, Quicktime, etc.1 The video should showcase core functionality of the data warehouse and ETL strategy. It should also show the execution of 2 or more of the BI queries discussed in the report. The voice commentary should explain key aspects of what is being shown, and noteworthy implications as well as interesting / surprising results or outcomes.
Some Guidance and Storage Requirements
Full freedom on the design and implementation methodology is given to the student. Students are free to use any data warehousing tool set discussed in the course, or otherwise.
The final project implementation should be stored until the module results are announced after the end of the semester. You may be called upon to demonstrate your solution if further details or evidence are required to support your paper and/or video presentation. It is your responsibility to ensure appropriate storage and backup for your project. Failure to do so, if demonstration is required, could negatively impact the project grade.