OW2con'15

Companies use only little over 10% of the owned data for analytical purposes, in order to monitor and improve their performance and decision making. On the other hand, an effective business strategy is based on the federation of data coming from inside the walls of your business (e.g. DWH, ERP, CRM, emails) and on their integration with external data sources (e.g. social networks, click streams). Businesses can turn data into information, also exploring people's attitudes, interests and opinions on products and services, in order to manage risks, anticipate customer needs and forecast wider market trends. This talk will show you how, thanks to an open approach and the integration of open source products, the Hadoop ecosystem offers effective solutions to achieve this objective. What's more, data scientist techniques will be applied to perform data mining algorithms using various open source tools (e.g. Spark MLLib, Mahout, R) to build predictive analytics. A real use case will support the presentation, showing how these technologies provide a seamless user experience, not necessarily requiring in-depth technical knowledge. We’ll use the charm of SpagoBI in a cloud environment designed with Canonical-Juju an open source, universal model for service oriented architecture and service oriented deployments.