BEGIN:VCALENDAR
PRODID:-//hacksw/handcal//NONSGML v1.0//EN
METHOD:PUBLISH
BEGIN:VEVENT
DESCRIPTION:Click for Latest Location Information: http://semtechbizsf2012.semanticweb.com/sessionPop.cfm?confid=65&proposalid=4670\nThis discussion centers on the application of semantic technology to advanced business analytics. Traditionally, predictive analytics centers on statistical analysis of amorphous raw data collected during complex business processes. Though such analysis is effective, many obstacles remain. Results of purely statistical methods are marked by associations that do not imply causation leading to spurious conclusions. Traditional methods of dimensionality reduction, such as principal component analysis and factor analysis, explain the output variations in terms of combinations of predictor variables that are difficult to interpret. CTG has developed innovative applications that employ hybrid architectures involving semantic and traditional technologies, resulting in higher quality outcome analytics in manufacturing and healthcare settings. Here we discuss how utilization of an upper level ontology, federated domain ontologies, and inference rules to filter input variables has proven valuable for linking data to relevant aspects of business processes thus allowing more focused and fruitful predictive analytics.\nThe goal of predictive analytics is to proactively assess processes in order to influence outcomes.\nMany interesting and important processes are marked by a high degree of variability.\nTraditional statistical analysis shows associative rather than causative relationships thus allowing spurious conclusions .\nSemantic technology allows automated organization and interpretation of raw data.\nHybrid semantic and traditional architectures have proven value in enhancing predictive analytics.
DTSTART:20120605T163000
SUMMARY:Semantically Enhancing Advanced Business Analytics: A Success Story
DTEND:20120605T165959
LOCATION: See Description
END:VEVENT
END:VCALENDAR