ICT2014: DISIT Lab, smart cloud knowledge modeling, and reasoning

DISIT Lab – one of the most active ICT labs of the University of Florence – successfully developed a number of International RnD/RIA/IA projects as well as solutions grounded on available DISIT tools for specified TRL – Publications: http://www.disit.org/5487 Tools: http://www.disit.org/5489 • Research areas: – big data, high performance distributed systems, data mining and understanding – semantic models and computing, knowledge mining and representation, ontology modelling, – artificial intelligence, natural language processing, • Techniques: data analytic, clustering, indexing and search, link discovering, regression, holistic regression, machine learning, prediction, inference, deduction, recognition, disambiguation. • DISIT solutions for: user behaviour analysis, recommendation, multilingual and cross media indexing, user and collective profiling, indoor/outdoor navigation, media synchronisation, matchmaking, audio transcoding, decision support, sentient and autonomous agents and tools, open data, linked open data. Recent Track Record on cloud • iCaro Cloud a smart cloud project on the verification of consistency and on smart strategies on cloud (6.4 Meuro): – http://www.disit.org/5604 – Cloud Monitoring – Cloud Interoperability – Cloud verification and validation • Linked Open Graph service: SPARQL/ Linked Data navigator open service and reasoner engine – http://LOG.disit.org Smart Cloud Model & Engine • Cloud for business acceleration of SMEs (developed in ICARO project) • Cloud Ontology and Knowledge base – ontology modeling cloud resources at level of IaaS, PaaS, and SaaS, SLA of multitier applications and deployments, monitoring data, supporting reselling, brokerage, etc. • Smart Cloud Engine – for automated provisioning and verification of service composition and deploy: a set of tools for reasoning about cloud status taking into account the cloud status and evolution via the Knowledge Base. – The intelligence on smart cloud is enforced by means of a set of algorithms to perform: • detection and prediction of critical conditions, • verification and validation of configurations (feasibility in terms of consistency and completeness and taking into account present and possibly available resources), • unexpected correlations about facts on cloud evolution, estimation of slack, automated verification of completeness and consistencies, etc. • Monitoring and Reporting tools – Estimation and production of high level metrics based on low level metrics, – generation of graphic and data results via services for reselling portals and cloud customers. Smart Cloud Ontology • Model different aspects of Cloud: – The infrastructure (Host Machines, Virtual Machines, Network, Storage, etc.) – The services available (database, filesystems, application servers, balancers, mail server, etc.) mapped to VMs, with monitoring aspects – The applications (built using services) with specific metrics and SLAs – The business