Context Addict Presentation

Context-ADDICT

Context -ADDICTContext -ADDICT

Politecnico di MilanoPolitecnico di Milano

Context-Aware Data Design, Integration, Customization, Tailoring

Context-ADDICT

Motivations and scenariosMotivations and scenarios

• Disparate, heterogeneous, independent Data Sources

• Semantic schema integration

• Context-aware information filtering: Data Tailoring

• Common, integrated, semantic access to data

• Issues: mobility, data transiency

• Multiple scenarios: system adaptability • < add your favourite buzz-word here >

Context-ADDICT

Tasks and ChallengesTasks and Challenges

Tasks:Tasks:

• Data Source Discovery (later)

• Lightweight (Semi)Automatic Data Integration

• (Semi)Automatic Semantic Extraction

• Context-Aware Data Filtering (focus)

• Semantic Distributed Query processing

Challenges:Challenges:

• Data Sources: heterogeneous, transient, mobile, unknown at design time

• User Mobility

• Multiple scenarios: system flexibility and adaptability

• Need for high automatism

• User Device Constraints (small portable devices)

Context-ADDICT

Overall System ArchitectureOverall System Architecture

Context-ADDICT

Models viewModels view

Context-ADDICT

Data TailoringData Tailoring

Data Tailoring, based on the Data Tailoring, based on the Dimension Tree InstantiationDimension Tree Instantiation::

• Schema Tailoring

• Instance Tailoring

Context-ADDICT

Data IntegrationData Integration

Domain ontology - Data source integration:

Standard Ontology mapping functionalities

Lightweight, automatic processing (mobile user’s device)

Automatic inconsistencies resolution

Context-ADDICT

Semantic ExtractionSemantic Extraction

Data Source Ontology:

• Semantic Extraction: data abstract model + storage model

• Supports the query processing

• Models isolation (different models can be used separately)

Context-ADDICT

Query AnsweringQuery Answering

Query Answering:

• Choose an ontology query language (SPARQL, OWL-QL)

• Query decomposition

• Query translation

• Data Fusion

• Query Optimization

Context-ADDICT

Context-ADDICT projects/thesisContext-ADDICT projects/thesis

We will managed area­based meeting/presentation:­ Ontology Mapping­ Semantic Extraction

­XML, Relational, Web(crawler)­ Ontology Tailoring­ Query Answering

Are you interested? (Please rise you hand when asked)

We will post the information about the meetings here:feed://www.elet.polimi.it/upload/curino/NEWS/rss.xmlhttp://www.elet.polimi.it/upload/curino/NEWS/NEWS.html

Context-ADDICT

Context-ADDICT projectsContext-ADDICT projects­ Dimension Tree + tailoring

­ ER tool integration­ X­SOM:

­ matching modules (neighborhood, subclass, probabilistic, H­MATCH integration)­ Protégé plugin and standalone

­ Relational Integration­ use CLIO (or similar) + automatic feeding by domain ontology

­ Query Answering­ query language selection (expressivity & al)­ automatic wrapper generation for Relational and XML

­ XML2OWL­ look at the XSLT based approach and enrich it...

­ Relational2OWL­ advanced features on ER generalization­ Plugin GUI

­ Ontology Extraction­ semantic completeness + labelling vs querying 

­ Web 2 OWL­ ontology extraction from web sources

QuickTime and aTIFF (LZW) decompressor

are needed to see this picture.

Context-ADDICT

Context-ADDICT teamContext-ADDICT team

Prof.ssa Cristiana Bolchini ([email protected])

Prof. Fabio A. Schreiber ([email protected])

Prof.ssa Letizia Tanca ([email protected])

Dott.ssa Elisa Quintarelli ([email protected])

Dott.ssa Rosalba Rossato ([email protected])

Ing. Carlo A. Curino ([email protected])

Ing. Antonio Penta ([email protected])

Ing. Giorgio Orsi ([email protected])

Context-ADDICT

ConclusionsConclusions

These projects are part of our research so are:

Limited, Challenging, Unique,

Work-intensive, Team-managed

If you want a project you are welcome on board, please contact:

[email protected]

[email protected]

Otherwise I’m sorry you have just lost the most challenging and exciting chance you have had in your life!