Data Mining Tutorial -- Classification Column

Classification Column

In the "Sales" table, suppose that you want to study the seasonal effect on sales. To do this, you need to classify the "Month" column into 4 categories; Winter, Spring, Summer and Fall. This can be accomplished by creating a new Classification column that has 4 values and 4 criteria as follows:

Value	Criteria
Winter	Is one of ( December, January and February)
Spring	Is one of ( March, April and May )
Summer	Is one of ( June, July and August)
Fall	Is one of (September, October and November)

Click on the first tab "Sales Data" in order to work with the Sales data view. Make sure that no filters are set on the view. To release all filters press on the

button, the green light should be off.

1. Place the cursor on the column " Month" and press the button on the "Sales" view. The Add / Edit Virtual Columns dialog is displayed. Type the new column name "Seasons " in the "Select / Add column" edit box. Press on the "Classification Column" button. The Enter Value - Criteria dialog is displayed.

2. The Enter Value - Criteria dialog allows you to construct a number of criteria - value combination. On the "Value" column type in the new four categories: Winter, Spring, Summer and Fall.


		5. Click on the drop-down box to select the operator " is one of". Select the values "December, January and February" from the list and press OK. The Enter Criteria dialog will show the criteria " is one of (December, January and February)" in the cell next to the "Month". Press OK to return back to the Enter Value - Criteria dialog. 6. Repeat steps 3 to 5 to construct the other three criteria. 7. Press OK. SuperQuery will create the new column "Seasons" and add it to the view before the "Month" column. The new column will be given a cyan color indicating that the column is a calculated column (Virtual Column). Immediately, you will notice that most sales were done in the Winter. Press the filter button when the cursor is placed on Winter, and move to the column "PRODUCT"; you will notice from the Reps Graph pane that both the Jacket and Shirt products are selling the most. Try to scroll in the different columns, set filters, and study the data more

		Chapter 1 -- Getting Started in SuperQuery Chapter 2 -- Organizing your Data Chapter 3 -- Querying the Database (Filter Column) (Classification Column) Chapter 4 -- Discover Facts and Rules (Office edition) (Discovery edition) Chapter 5 -- Summarize your data


		3. To create the first condition for Winter; "Month is one of (December, January and February)" Double click on the cyan cell next to the first value "Winter". The Enter Criteria dialog is displayed.


		4. Double click on the cell next to the "Month". The Edit Criterion dialog is displayed.