Introduction

An IBM industry model solution is a comprehensive set of industry-specific predesigned models that form the basis of a business and software solution, optimized for business challenges in a particular sector. Domain areas include data warehousing, business intelligence, business process management, service-oriented architecture, business terminology, and business glossary templates.

The IBM Industry Models solutions cover a range of industries that include banking, health care, retail, and telecommunications. The IBM Industry Models solutions provide you with an extensive and extensible data model for your industry sector. Use the logical data model as provided by IBM to build a physical data model that is customized for your reporting requirements, and then deploy and populate an IBM InfoSphere®
BigInsights™environment.

This paper shows how you can deploy the IBM Banking Data Warehouse (BDW) solution to BigInsights. This paper introduces the logical data model concept and then focuses on what you must do to transform a non-vendor-specific logical data model into a production-ready BigInsights BigSQL schema. The key steps of the deployment process are as follows:

1. Create a physical data model that is a subset of the supplied logical data model.
2. Refine the physical data model to reflect your reporting and analytics needs.
3. Create the physical schema from the physical data model.
4. Populate the test environment, to further optimize the physical design to reflect the anticipated query, ingest, and maintenance workload.

The example used in this paper is the Involved Party coverage within the BDW. Although BigInsights includes a SQL interface (BigSQL), there are unique considerations when deploying to BigInsights. This paper explains how to translate reporting needs into BigInsights design decisions as part of the transformation process.

InfoSphere BigInsight Summary

InfoSphere® BigInsights™ is a software platform for discovering, and visualizing data from disparate sources. You can use this software to help process and analyze the volume and variety of data that continually enters your organization. BigInsights is built on the scalable Apache Hadoop open-source framework that runs on commonly available, low-cost hardware. By using BigInsights, users can extract new insights from data to enhance their knowledge of your business.
BigInsights includes text analytics for analyzing large volumes of text, which often contains extraneous data ("noise") that is not important to your business, to gain insights into unconventional data. You can incorporate these capabilities into your IT infrastructure to complement and extend existing analytic capabilities by augmenting your data warehouse with these insights. You can use BigInsights as an engine to store, filter, analyze, and transform incoming data. You can then extract business insights and important summary data and then load them into your data warehouse. You can also use BigInsights as a query-ready archival system for your data warehouse to quickly access data that is typically archived. You can offload data to BigInsights, where you can query it at any time. This implementation saves time and resources by integrating into your existing architecture.

BigSheets is a browser-based analytic tool that is included in the BigInsights console. You can use BigSheets to collect data from multiple sources, format it, and explore it by using a spreadsheet-like interface. You can also apply visualizations such as tag clouds, bar charts, maps, and pie charts to provide consumable output that highlights relationships and distill insights from previously disconnected data.

Of particular relevance to this paper is BigSQL. BigSQL is a software layer that you can use to create tables and query data in BigInsights by using familiar SQL statements. BigSQL uses standard SQL syntax and, in some cases, SQL extensions that IBM created to make it easy to exploit certain Hadoop-based technologies. The BigSQL query engine supports joins, unions, grouping, common table expressions, and other familiar SQL expressions. Depending on the nature of a query, the data volumes, and other factors, BigSQL can use Hadoop's MapReduce framework to process various query tasks in parallel or run a query locally within the BigSQL server on a single node whichever is the most appropriate for the query.

Download the full report for more on deploying the IBM Banking Data Warehouse to IBM InfoSphere BigInsights.
Download the report to get started!

#Db2