Supporting semantic web technologies in the pervasive service ecosystems middleware
Paolo Contessi
Semantic Web technologies are strategic in order to fulfill the openness requirement of Self-Aware Pervasive Service Ecosys-tems. In fact they provide agents with the ability to cope with distributed data, using RDF to represent information, ontologies to describe relations between concepts from any domain (e.g. equivalence, specialization/extension, and so on) and reasoners to extract implicit knowledge. The aim of this thesis is to study these technologies and design an extension of a pervasive service ecosystems middleware ca-pable of exploiting semantic power, and deepening performance implications.
Mirko Viroli (Supervisor)
Danilo Pianini (Co-supervisors)
Main Goals
- Acquire knowledge on technologies to be exploited: RDF, SPARQL/SPARUL, OWL/RDFSchema, Desanti's LSA-space (done)
- Analysis and development of Ontologies support in SAPERE (done)
Semantic Matching(on-going, delegated to Gabriele Manduchi)- Platform performance profiling (done)
- Implementation of a Eco-laws->SPARQL/
SPARULstatements translation (optional) (suspended, TR.WP1.2011.06 partially supported)
Downloads and Links
- Deliverable from WP1 D1.1 (available on SAPERE project Wiki)
- Repository (private, at least read rights are required)
- W3C RDF primer
- Semantic Web Programming (book used as reference)
- Jena Tutorial (and aceyogi's version, needs permissions)
- Maven and relatives references (from Apache and Sonatype)
- Queries Experimenter (utility for SPARQL/SPARUL queries testing on an RDF Graph Store, produced during development)
Activities
Detailed phases and sub-goals, which are concerned with the development of the thesis, will be weekly reported in the following.- Week 1 (11/01/2012 - 15/01/2012)
- Orienteering and setup
- Start working on Main Goal (1)
- Revising process-algebras and operational semantics theory from LMC LM
- Week 2 (16/01/2012 - 22/01/2012)
- Studying Desanti's thesis
- Analysing saperenode code, cloned from remote repository on BitBucket
- Trying to determine basic interfaces and how to insert new features
- Starting definition of a workflow
- Week 3 (23/01/2012 - 29/01/2012)
- Continuing from Week 2
- Collecting main concepts from Semantic Web Technologies
- Week 4 (30/01/2012 - 05/02/2012)
- Continuing from Week 3
- Week 5 (06/02/2012 - 12/02/2012)
- Continuing from Week 3
- Designing the extension of saperenode
- Week 6 (13/02/2012 - 19/02/2012)
- Continuing from previous weeks
- Working on LSA model
- Week 7 (05/03/2012 - 11/03/2012)
- Forking SAPERE OSGi platform prototype (from UNIMORE)
- Week 8 (12/03/2012 - 18/03/2012)
- Formalizing RDF-based LSA model
- Week 9 (19/03/2012 - 25/03/2012)
- Providing first implementation of LSA-space and SAPERE (LSA) model
- Test and validation
- Week 10 (02/04/2012 - 08/04/2012)
- Configuring and running integration tests
- Integration with Ecolaws
- Week 10 (09/04/2012 - 15/04/2012)
- Redefining Ecolaw model, in order to fit with new requirements
- Week 11 (16/04/2012 - 22/04/2012)
- Continuing from previous week
- Approaching to Ecolaw2SPARQL serialization
- Week 12 (23/04/2012 - 29/04/2012)
- Ecolaw2SPARQL serialization
- Revisioning work against new requirements
- Week 12 (30/04/2012 - 06/05/2012)
- Fighting with Pellet + Jena integration in an OSGi context
- Considering possible solution for new LSA specifications
- Integrating developed LSA-space and models into the rest of the middleware
- Week 13 (07/05/2012 - 13/05/2012)
- Integrating developed LSA-space and models into the rest of the middleware
- Week 14 (14/05/2012 - 20/05/2012)
- Testing integration
- Trying an Android porting
- Week 15 (21/05/2012 - 27/05/2012)
- Testing integration
- Checking features support
- Implementing Multi-Level LSA support
- Testing the overall system against new requirements
- Android porting suspended
- Week 16 (28/05/2012 - 03/06/2012)
- Defining a (micro-)benchmark scenario
- Debugging
- Week 17 (4/06/2012 - 10/06/2012)
- SAPERE meeting in Reggio Emilia
- Implementing (micro-)benchmarks
- Week 18 (11/06/2012 - 17/06/2012)
- Running (micro-)benchmarks and collecting results
- Analysing results
- Defining demo scenario
- Week 19 (18/06/2012 - 24/06/2012)
- Refining demo scenario
- Implementing DIFFUSE support (relocator)
- Defining demo scenario: Sensor+Aggregation+Diffusion
- Trying to determine max throughput without ReactionManager starvation
- Week 20 (25/06/2012 - 01/07/2012)
- Continuing from Week 19
- Week 21 (02/07/2012 - 08/07/2012)
- Continuing from Week 20
- Testing platform scalability
- Week 22 (09/07/2012 - 15/07/2012)
- Analyzing performance data
- Correcting diffusion eco-law
- Deploying release
- Week 23 (16/07/2012 - 22/07/2012)
- Writing thesis
- Week 24 (23/07/2012 - 26/07/2012)
- Completing subscription
- Creating slideshow
