I am currently working on a chapter about in-database analytics for a new book which will published early next year. During my research it has been interesting to note all the different specialised engines that have evolved to solve specific business problems. However, using isolated stand-alone analytical engines creates problems for business users because they end up building numerous data-silos and sharing data becomes a total nightmare. The Oracle Database supports a wide range of analytical features: data mining, text mining, spatial, graph, statistics, multidimensional (OLAP) etc etc. By creating a integrated analytical platform, Oracle allows developers and business users to create analytical mashups that they can then embed in applications, reports and dashboards. So far so good. Oracle has a great story for analytics: single, complete, integrated analytical platform.
So Oracle has worked hard to help customers avoid building analytical silos but now there is a new problem on the horizon: visualization silos. There is an interesting article here http://www.netmagazine.com/features/top-20-data-visualisation-tools which explores some of the many data visualisation tools that you can use to display unstructured and structured data. The increasing interest in big data is driving the need to have more powerful visualisation tools beyond the simple cross tabular data reports and line/bar charts. This article contains a long list of tools and products. What worries me is that there is very little overlap between many of these products and tools., which means that there is huge potential here to build visualisation silos. Now all we have done is take the bad old days of building analytical silos and move it to our visualisation layer which is going to cause the same problems as before and mean lost opportunities and frustrated business users.
For example, what if I need to show the results of a statistical analysis alongside my geographical map? Looking at the list of products and tools in the article above you would probably have to start a major IT project to glue together bits from different toolkits.
So while the Oracle DB development team has being doing a great job creating a single integrated analytical platform the Oracle BI team have been doing the same thing at the visualisation layer.
There is an excellent webcast next week that demonstrates this very nicely. The title of the webcast is "How to use Oracle’s Spatial and BI Tools for location-aware predictive analytics" and is scheduled to start at 2:00pm EST on November 29 (click here to register: https://www2.gotomeeting.com/register/764677554). During this webinar, Oracle experts will show how to perform predictive analytics, network analytics and spatial analysis – combined together, in real-world scenarios. Demos will include evaluating airline on-time performance and retail establishment performance.
A perfect example of an analytical mash-up (spatial + data mining) alongside a visualisation mash-up and it highlights the key advantage of Oracle:
- all the analytics running inside the Oracle Database
- all the visualisations being managed by Oracle BI
No data silos and no visualisation silos. Oracle: Better Together!