Combined Ontology-Driven and Machine Learning Approach to Monitoring of Building Energy Consumption

A number of studies within the building domain haveindependently considered knowledge representation andreasoning (KRR) and machine learning (ML). This paperexplores opportunities for how these artificial intelligence(AI) technologies can be combined to provide synergicassistance in intelligent building systems. A case study isused to demonstrate the construction of semantic knowledgefrom weather and utility ontologies. That knowledgeis used as a semantically annotated training set in clusteringalgorithms that identify consumption patterns, and theresults of clustering are stored in ontologies for furtherinference. This paper presents a roadmap to intelligentbuilding monitoring techniques that uses both historicaldata and the underlying semantic knowledge.