The project focuses on performing data warehousing techniques on real crime datasets to answer some business queries.
- PowerBI file
- PDF report generated from PowerBI
- Data cleaning/ETL script
- SQL script and cleaned CSV files
- Solution project file (and its folder) of the SSDT analysis service multi-dimensional project
- PDF file discribing the association rule mining process and results.
- Consider a few business questions my data warehouse could help answer.
- Draw a StarNet with the aim to identify the dimensions and concept hierarchies for each dimension. This should be based on the lowest information level I can access.
- Use the StarNet footprints to illustrate how my design can answer business queries.
- Once the StarNet diagram is completed, draw it using a tool or software - in this project, diagram.net was used, and paste it into PowerBI
- Implement a star schema using SQL Server Management Studio (SSMS) and paste the database ER diagram generated by SSMS onto Power BI Dashboard.
- Load the data from the CSV files to populate the tables by using the SQL code command
- Use SQL Server Data Tools to build a multi-dimensional analysis service solution, with a cube designed to answer my business queries. Make sure the concept hierarchies match StarNet design. Paste the cube diagram to your Power BI Dashboard.
- Use Power BI to visualise the data returned from the business queries.
The Association Rule Mining is applied in this crime dataset.
- Process the dataset into a case table and a nested table.
- Explain the top k rules (according to importance or probability) that have the "crime" type (or other suitable columns) on the right-hand side, where k>=1.
- Explain the meaning of the k rules in plain English.