The Internet of Everything and the Semantic Web can be joined in the Semantic Web of Everything (SWoE), giving more intelligence to pervasive systems at the network edge. To that end, knowledge representation and reasoning capabilities must be stretched towards the edge computing paradigm, extending support to resource-constrained embedded devices. This thesis delivers the following contributions towards the achievement of the SWoE vision, discussing design details and providing results of an experimental evaluation campaign: Tiny-ME, a novel cross-platform automated reasoner and matchmaking engine; Cowl, a knowledge representation library targeted at embedded devices; evOWLuator, a benchmark framework that allows the appraisal of energy consumption, CPU and memory usage, while also being able to carry out inference tasks on remote devices.
L’Internet of Everything ed il Semantic Web possono essere fusi nel Semantic Web of Everything (SWoE), fornendo maggiore intelligenza ai sistemi pervasivi al bordo della rete. A tal fine, capacità di rappresentazione della conoscenza e ragionamento automatico devono essere estese all’edge computing garantendo il supporto a dispositivi integrati con risorse di calcolo fortemente limitate. La tesi fornisce i seguenti contributi verso la realizzazione della visione SWoE, discutendone dettagli progettuali e presentando i risultati di una campagna di valutazione sperimentale: Tiny-ME, ragionatore automatico e motore di matchmaking multi-piattaforma; Cowl, libreria per la rappresentazione della conoscenza orientata ai dispositivi integrati; evOWLuator, framework di benchmark che consente il monitoraggio del consumo energetico in aggiunta all’uso di CPU e di memoria e l’esecuzione di task di inferenza su dispositivi remoti.
RAGIONAMENTO AUTOMATICO PER L'EDGE COMPUTING: DESIGN, PIATTAFORME E VALUTAZIONE SPERIMENTALE
BILENCHI, IVANO
2019/2020
Abstract
The Internet of Everything and the Semantic Web can be joined in the Semantic Web of Everything (SWoE), giving more intelligence to pervasive systems at the network edge. To that end, knowledge representation and reasoning capabilities must be stretched towards the edge computing paradigm, extending support to resource-constrained embedded devices. This thesis delivers the following contributions towards the achievement of the SWoE vision, discussing design details and providing results of an experimental evaluation campaign: Tiny-ME, a novel cross-platform automated reasoner and matchmaking engine; Cowl, a knowledge representation library targeted at embedded devices; evOWLuator, a benchmark framework that allows the appraisal of energy consumption, CPU and memory usage, while also being able to carry out inference tasks on remote devices.I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14253/106949