Semantic Modeling for Automated Decisions

Semantic models provide a good data basis for intelligent systems from decision automation to data analysis. They complement black-box statistical approaches with an analytic body of knowledge – this way they can add transparency and intervention options to machine learning, optimization or text analysis methods.

This session will discuss semantic modeling practices for automated decisions from a variety of industry projects:

How do we integrate learning's from statistical analyses in semantic model and, conversely, use the model as background for analysis?
How much do we have to model and how do we know when to stop?
How do we deal with uncertain and incomplete knowledge?
How do we engage the domain experts in the knowledge engineering?

Project examples include:

Compatibility of products based on their characteristics
Job and Business Matching
Prediction of the success probability of legal claims
Making sense of sensor data in traffic analysis
Resource optimization constrained by complex business rules

Over the last years, Klaus Reichenberger developed together with his team solutions for more than 40 clients - based on intelligent views' semantic graph database k-infinity. He has over 15 years of experience in the field of semantic technology and graph-based representation of knowledge. He is author of numerous publications and a frequent speaker on semantic modelling and its applications in the enterprise. Prior to founding intelligent views, Klaus Reichenberger worked as a researcher for the Fraunhofer-Gesellschaft bringing together knowledge representation and visualization techniques.