Karachi

点击翻译按钮获取中文摘要

The key aim of the ASTERI project is to demonstrate the automated creation of a Star Schemarelational data warehouse representation of the database for an On-line Transaction Processing(OLTP) system. Database structures for OLTP systems are optimised for data storage and soare poorly structured for obtaining analytics-oriented information. Star Schema-basedsolutions are optimised for data retrieval i.e. the creation of analytics. Ease of access toanalytics by a wide range of users is of increasing commercial importance.The innovation in the ASTERI project is based upon the application of:a) Machine learning techniques to produce the transformation from the original OLTPdatabase format to the relational database form used in the data warehouse. The data engineeris then used to confirm a transformation as opposed to its creation.b) Event-driven visualisation using HTML5 based data presentation. This includes addressingefficient and responsive visualization of 100s of thousands of data points. The associatedreports will then reflect the latest data available.While the significance of Star Schema to data warehouse representation has been well knownfor the past 20 years, it is only in the past 5 years that research into the use of ‘MachineLearning’ to create a Star Schema representation has shown the potential of the approach.With these new capabilities Rosslyn Analytics (RAL) can substantially reduce the effort andcost to engage new customers and can rapidly create new sector database schemas enabling itto expand into new markets and gain market share with smaller organisations. RAL alsointends to leverage the new tools to expand its emerging OEM sales to major global dataanalysis service providers such as Xerox, Pitney Bowes and Qlikview to be used as their dataanalysis service engine. In addition, the automated tools will permit increased sales bydevelopment partners who use the RAL RAPid toolset to deliver value for their own clients