Dynamic Context Aware Workflow Composition in Internet of Things

Abstract

Internet of Things (IoT) is rapidly expanding its reach into a multitude of real-world newlinesystems. Increased adoption has led to a sudden spike in the number and variants of device, newlineresulting in interoperability issues. Existing solutions typically address communication newlineinteroperability by defining protocols or handle it at the application level by a software newlineplatform. At an application level, interoperability is about harnessing the meaning newline(semantics) of the data which opens up more possibilities to make the machine to machine newlinecommunication seamless. Ontologies formalize the domain knowledge and enable machines to newlineinterpret the information without human assistance. Across IoT domains (Healthcare, newlineLogistics, Energy, Home etc.) there are more than 800 documented ontologies that exist as of newlinedate, leading to huge fragmentation of domain knowledge. Hence, the need of the hour is to newlineconsolidate the domain knowledge, by matching concepts, relations and align them with a newlinereference ontology. A reference ontology, Comprehensive Ontology for Internet of Things newline(COIoT) and ontology matching algorithm Multi Ontology Matching Generative Adversarial newlineNetworks for Internet of Things (MOMGANI) are introduced to achieve the objective of newlineconsolidation of domain knowledge. COIoT, defines key concepts such as sensing, actuating, newlinelife-cycle, service, policy, context and monitoring carefully avoiding redefining concepts still newlinebeing agnostic to a domain. Ontology matching and alignment algorithm MOMGANI, has newlineproduced comparable performance to that of a hand-coded ontology clocking precision, recall, newlineF-measure of 0.91, 0.80 and 0.87 respectively which is very close to the baseline matching newline(hand-coded) performance. Several aligned ontologies, lead to development of a unified newlineknowledge-base that can promote large scale operation of devices within the same domain or newlineacross domains (e.g., Smart City). In smart environments, it is critical for processes to adapt on newlinethe basis of the operating context. Business Process Modeling Notation-Things Extension newline(BPMN-TE) enables expression of the operating environment constraints, resources, humans, newlineSpatial, Temporal and Environmental factors as first-class constructs at design time. The newlineoperating context and the results of the aligned ontologies codified into a knowledge-base newlineenables dynamic execution of the automated process thus enabling seamless machine to newlinemachine communication. newline