Oracle Visual Analyzer Project: Data-Mashup Demo
Oracle Visual Analyzer (VA) is part of the OBIEE 12C release that might be available in early- to mid-2016. When it arrives, VA will provide business intelligence (BI) users with visualization tools to make their analysis more personalized and more visually appealing.
Easy search and find
The first thing you’ll notice when you log into VA is the new interface and the ease of navigation. One feature that stands out to me is the “Finder,” that lets users quickly search through BI content. The “Finder” reminds me of the spotlight search on MacBook and iPhone in that the content search starts as soon as you start typing.
Another great feature that will put this tool on the BI map is the ability it gives users to integrate owned and supplied data with existing OBIEE subject areas, also known as Data-Mashup. Data-Mashup allows the BI user to expand the existing subject-area data model to include new factual and/or measured information that he/she own and join it to an existing OBIEE subject area.
In the example below, I go into more detail to demonstrate how easy this is to do in VA, with a demo of the “Data-Mashup” using an external Excel spreadsheet.
VA is accessed through a standalone URL ending in “/VA.”
Click on the VA icon to start a new Data-Mashup project.
Add the first data source needed for the Data-Mashup. In this case, we will select a subject area that relates to the Excel sheet provided and add it to a project.
The subject area is then added to the view.
Now, let’s add another data source for which we would like to join to the “Procurement and Spend” subject area, in this case, the data source is an Excel spreadsheet.
Click on “Upload File” on the screen, then select “File” to locate your Excel analysis.
Once the Excel analysis is added, you will see all the columns laid out as below, and
VA will take the column headers from the Excel analysis and perform the following:
1. Add an aggregation rule to the column if it is a numeric column. In cases where a numeric is not a measured column, in this example, “Department Code,” the user can switch it from **“Add Measure” to ** “Match With” if the column exists in the subject area
or ** “Add Attribute.”
3. Match column headers with columns in the subject area. In cases where a match cannot be made between the Excel analysis and the subject area, the user can “Add Attribute” which will add the column as a data source in the project.
In this step, the user should review all the columns added from Excel analysis and ensure that fact and dimensional columns have been/are identified as so, new columns are added to the project if needed, and the proper aggregation rules are applied on the measures. The user, at this step, has the ability to do the following:
Match columns by name to the added subject area if VA did not already do that.
**Match With: Matches by name.
**Add Attribute: Will add the column to the project as a new attribute.
**Add Measure: Will add the column as a measure attribute and allow the user to apply an aggregation rule.
Once the import is complete, you should see two data sources added to the project: one being the subject area and the other being the Excel analysis.
Save the project.
Now let’s create an analysis using the two data sources that we added to this project:
1. Bring the “IT Leader” column that exists in the Excel analysis but not in the subject area.
2. Bring columns that we know exist in both data sources, like “Department Code,” so we can ensure that the data lines up between the two sources.
3. Finally, let’s add a measure column, in this case, “AP Amount.”
On the next installment of this blog, we will be talking about the new archiving method of OBIEE components across environments, called BI Application Archive (BAR) or the “Multi-Tenancy Service” whichever the beta program image allows us to use.
About the beta program: I ran into a few issues when putting this demo together. I tried to hide them as best as I could. However, because I am part of the Oracle beta program, any issues that I encounter during demos are reported to Oracle so that fixes can be developed.