by Madan Thota

Social data has become more significant in business intelligence.  This article profiles how  Oracle Endeca Information Discovery can read data from social sites and how we can combine this data with other sources.  This 4-step process shows how you can get started using the Oracle Endeca Information Discovery Application for Social Media Analytics. 

4 Step Process for Any Social Media Discovery Project

Step 1 - Requirements

Identify key KPIs and metrics you would like to analyze:

• Identify quantitative metrics such as counts, posts and followers.
• Identify qualitative information like comments, feedbacks and sentiment.
• Plan what external data to merge with, and how to combine social media data with existing information (like revenues).

Step 2 - Extract, Refine & Load

• Build integrator graphs and load data. Based on source systems, make the decision to use EID features like http connector, web crawlers, Java Database Connectivity (JDBC) connectors.
• Social data can be retrieved in the format of JavaScript Object Notation (JSON) files.
• Load this data with JavaScript Object Notation (JSON) Reader components or through provisioning services. 
• The data can be combined with other sources information though graphs.
• All relevant data can be loaded into same data store.

Step 3 - Visualize and Analyze

• Build the studio application. 
• Once the data is loaded Endeca Data Store, start with Endeca Studio to build the dashboards. 
• The search and guide navigation features of Endeca allows users to filter and discover the data in depth. 
• Plan for visualizations like metric bars, graphs and tag clouds to make the analysis more meaningful.

Step 4 - Administration

• Perform administrative activities like security setup before the data or studio application is released to users.  

Madan Thota is a Director of Services at KPI Partners and works with the expert team within the KPI Partners Offshore Technology Center.  His professional focus is in the world of enterprise decision support, business intelligence, and data architecture.  Check out Madan's blog at KPIPartners.com.