Data mining model management to support real-time business intelligence in service-oriented architectures
İsmail Arı, Jaja Li, Alex Kozlov, Mohamed Dekhil · 2011
Use of predictive models for making business-critical decisions is on the rise. However, serious challenges remain on managing data mining models and integrating them with business services using service-oriented architectures (SOA) to provide real-time Business Intelligence (BI). These challenges include model aging, management scalability, timely-communication among parties on model changes, semantic gap on interpreting models, and business process integration. We describe a data mining model management system that addresses these challenges to support sustainable and operationalized BI.