Sneak peek: Oracle Data Visualization Desktop
Our review and implementation guide
June 24, 2016 | By Chandrakumar Karunamoorthy
It’s not every day a Gartner update causes controversy. But when the IT advisory firm released its influential Magic Quadrant for business intelligence in February, the IT world was shocked to find most vendors had slipped from the leader quadrant—and Oracle had fallen off altogether.
What happened to cause this big shift? Gartner updated its model for evaluating business intelligence (BI) platforms to reflect a trend of data and analytics tools increasingly being used and purchased by departments outside IT.
Oracle’s powerful BI platform—which many enterprises use for a host of functions, from human resources to enterprise resource planning—is a product that requires heavy technical skills to implement, integrate and use.
Gartner’s shakeup left many IT buyers wondering if Oracle’s BI products are right for them to purchase or continue using. With the growing trend of self-service BI products that don’t require heavy IT involvement, Oracle’s model may indeed seem dated. While most large enterprises need a robust platform to link large quantities and types of data to produce the most powerful results, they’re also facing increasing internal demand for self-service capabilities.
Where does that leave enterprises considering investing in or maintaining their Oracle platforms? Aravind Darla, a senior software engineer with TEKsystems, thinks Oracle isn’t down for the count. “For enterprises dealing with large amounts of data—and who care about speed and performance—I still recommend Oracle. It’s a more complete solution.”
“What people don’t always think about when they’re evaluating vendors is support,” adds Darla. “Oracle has great support, great customer representatives to help you figure out how to actually use your tools.”
[Related: The 10-minute read on successful BI implementations]
Can DVD help Oracle compete with Tableau for self-service BI?
With the release of Data Visualization Desktop (DVD), Oracle may have found a way to compete with other intuitive self-service tools offered by competitors like Tableau and Qlik.
Our team of architects and developers had a chance to try the DVD tool and we found a lot to like:
- DVD is easy to implement for any type of user
- It provides stunning visualizations that allow users to explore patterns and outliers in their own way
- DVD is a single-user desktop application, with no remote server infrastructure
- The insights discovered are easy for users to package and share
The nitty gritty: How to implement Oracle Data Visualization Desktop
We found the DVD implementation to be refreshingly simple, and well within the capabilities of your average IT or business user. If you’re looking to implement DVD yourself, you can follow this step-by-step tutorial:
Under the hood, the Data Visualization Desktop tool is built on Visual Analyzer, present in OBIEE 12c and BICS. The installation includes a sample project, along with underlying data sets, and it can be used as a quick reference when we create our own visualizations. This illustrates the use of many features, such as forecasting, calculated columns, insights and stories.
When we launch the application, we are presented with the following home screen with the list of projects created including the sample project. A new Visual Analyzer (VA) project can be created by clicking the Create VA Project link on the left.
Before creating a project, we would need to create data sources we plan to research. We have three options: first, using Excel file as a data source is straightforward. Second, we can connect to an existing Oracle application and use Analyses/Reports as data sources. Third, we can import tables from a database.
Upon selecting the From a File option and navigating to the file and choosing it, we can preview the columns and sample data and click the Add Data Source button to add the file as a data source. The newly added data source appears in the list of data sources as shown below. We can now choose this data source when creating a new project and create visualizations and analyses.
Creating visualizations is as easy as drag and drop, and we created our first project (shown below) in minutes. In addition to the table and pivot table, the tool includes 20+ chart types like sunburst, radar, donut, scatter (cat.) and so on.
For adding a data source from Oracle Application, first we need to create a new connection to the Oracle application by specifying the URL, username and password as shown below. Under Data Sources we can click the Connections button to create a new connection. Once the connection is created, we can browse the Oracle application catalog and select any analysis/report as our data source.
If not present already, you may want to first create an analysis/report in Oracle Application with all the attributes and measures you would like to analyze using Data Visualization Tool, and import it as a data source. The data in the data source will reflect any filters or selection steps included in the BI analysis/report. A cached copy of the data source is created locally, and this data and metadata (column name and datatype) can be refreshed as needed.
All the attributes and measures present in the analysis/report created as a data source are available for creating visualizations in Data Elements pane in the project, as shown below.
To add more data sources to an existing project, we can right-click the Data Elements pane in the project and select the Add Data Source option to choose from available data sources already created, or we can create a brand new data source and add it to the existing project. The tool automatically matches the new data source with the existing one based on the common column. By clicking Source Diagram, we have an option to edit the join (created by the tool) manually to add more columns to the join, or to change the existing column in the join.
Once the data sources are created and added to the project, multiple visualizations/charts can be created by clicking the Add Visualization icon in the project toolbar. Creating a visualization is intuitive and easy. We can select multiple columns on the Data Elements pane and right-click and choose the visualization type from the list to create a new visualization. There is an option to create best visualization based on the columns selected. The visualization type can be changed after it is created as well, for example, from bar chart to pie chart.
Add Visualization button:
We can reposition the visualizations present in the canvas by dragging and dropping to a different position easily. The size of any visualization can be changed by dragging the edges. There are options to customize the color scheme, add a background image using free-form layout, change the color of a particular attribute/measure or a particular data point, create calculated columns (e.g. Profit Ratio = Profit / Sales * 100), and control the interactions between visualizations for users who appreciate more options and control.
For using built-in advanced analytics features like forecast, clusters and outliers, we need to install Advanced Analytics based on R distribution. This is not installed by default when we install the DVD tool, and it needs to be installed later. It can be installed in under 10 minutes by clicking the Install Advanced Analytics option located in the Oracle Programs folder. Once installed, it is as simple as right-clicking on any visualization and adding clusters, outliers, etc.
Oracle Programs folder:
The process of adding data sources from Database is similar to the second option mentioned above. We need to first create a new connection to the database and then create new data sources from tables in the database. Most of the leading databases are supported in the DVD tool shown below. Once the connection is established we can create as many data sources as we want from that database.
When creating data sources from a database, we have an option of either keeping the query mode live, meaning the system does not load the table and queries are sent live to the database, or auto mode, in which a local cache of the table is created and used when possible. While creating the data source there is an option to write SQL to join multiple tables and get the columns needed as just one data source as well.
Within a project we can add different data sources and join them by right-clicking on the Data Elements pane and clicking Add Data Source. This is useful when our data elements are present in more than one data source, for example, one data source for dimensional attributes and another for fact measures.
One of the key features is the ability to build stories and share them with others. We can capture insights, group them into stories and share them. Insights enable us to take a snapshot of the information we see in a visualization, and help us remember “aha” moments while we work with the data. Here’s how: In the Insights pane click the Add Insight link.
The Insight’s name and description can be edited to describe the information present in that Insight, by clicking the down arrow and selecting Edit. We can continue adding more than one insight to build a story about our exploration. The story builds in the Story Navigator (Story Navigator icon in Project toolbar) with each insight displayed as a blue circle as shown below.
To share the story, we can export the entire project as a .dva file that can be imported by other users. The export includes everything other users will need, like associated data sets, the connection string, connection credentials and stored data. There is an option to exclude connection credentials while exporting the project to make the importing user supply the connection password. To export a project, we would locate the project on the home page and click Options and select Export.
Have you used Oracle DVD? We’d love to hear about how enterprises are using it. Share your experiences in the comments.