ISWC '18 Doctoral Consortium Presentation

Ben De Meester, Ghent University – imec – IDLab, Ghent, Belgium

High Quality Schema and Data Transformations
for Linked Data Generation

Ben De Meester, ben.demeester@ugent.be, @Ben__DM

Supervisors: Anastasia Dimou and Ruben Verborgh

Ghent University – imec – IDLab, Ghent, Belgium

imec imec

Relevancy: Garbage In, Garbage Out

Geo errors
Janowicz et al. Moon Landing or Safari? A Study of Systematic Errors and Their Causes in Geographic Linked Data

Problem Statement

Computationally intensive

Related Work

Research Questions Applying validation on the generation process

  1. Can we provide a use-case independent and declarative
    Linked Data generation description?
    1. Can we declaratively define data transformations?
    2. Can we provide a complete generation description?
  2. Can we automatically validate a generation description
    based on the constraints for the resulting RDF graph?

Hypotheses: generation

Hypotheses: validation

  1. We can create a validation approach allowing custom entailment during validation that is comparable to the SotA w.r.t. functionality and speed.

Approach

id date 1 April 1st ‘89 ex:{ id } foaf:age { *age* }^^xsd:int Data transformation Validation + entailment Schema transformation ?s foaf:age ?age ?age :type xsd:int 1 2 3 4
  1. Reusable data transformations
  2. Aligned transformations
  3. Validation with custom entailment
  4. Automatic validation on generation process

Preliminary Results: Generation

Preliminary Results: Validation

Evaluation Plan Apply validation to the generation process

Reflections

High Quality Schema and Data Transformations
for Linked Data Generation

Ben De Meester, ben.demeester@ugent.be, @Ben__DM

Supervisors: Anastasia Dimou and Ruben Verborgh

Ghent University – imec – IDLab, Ghent, Belgium

imec imec
Ghent University imec