Skip to end of metadata
Go to start of metadata

Description

The course covers a sample worflow starting from non-RDF datasets, using OntoRefine to clean them up and transform them, loading the data using the SPARQL endpoint into a repository in GraphDB, adding structure to the data, running queries, using LOD resources.

The pre-requisites for the course are covered in the form of pre-class video materials & SPARQL exercises familiarizing the audience with semantic technologies (RDF, RDFS, OWL, SPARQL, graph databases, LOD, semantic data integration).

Attendees are also entitled to an hour long consultancy session which is used to address their specific problem and establish potential for future collaboration or services that Ontotext can offer.

Structure

The course consists of 4 components:

  • 2-3 hours of video materials on Semantic Technologies
  • 1-2 hours worth of SPARQL exercises (as a handout)
  • 4 hours live group training session in designing a simple PoC
  • 1 hour individual consultancy session

Video materials on Semantic Technologies

The pre-class video materials cover the following topics:

  • Definition & overview of Semantic Technologies
  • RDF
  • RDFS, OWL, other commonly used ontologies
  • SPARQL
  • Graph databases, triplestores, reasoning strategies
  • Semantic data integration
  • Use cases / reference projects
  • GraphDB Workbench overview
  • OntoRefine

SPARQL exercises

A sample dataset is handed out with a list of tasks to be completed on it in the form of SPARQL queries.

Designing a simple PoC live session

The 4-hour training is split in two 2-hour sessions with an hour break inbetween. Each 2-hour session has a short 10 minute break in the middle.

The live session consists of:

  • 5-10 minutes introduction & agenda
  • 1h15m - 1h30m demonstration of a sample workflow
  • 1h30m - 1h45m interactive group PoC design (similar workflow, different dataset)
  • 15-30 minutes Q&A session

The sample workflow consists of:

Step Companies (& countries) data Movie data
Using OntoRefine to clean up non-RDF data
facets, value transformations, splitting columns, creating new columns facets, value transformations, splitting columns, creating new columns
Transforming the data and loading it into a repository
using a SPARQL endpoint, insert query for loading data,
optional patterns, bind functions, SPIN RDF
using a SPARQL endpoint, insert query for loading data,
optional patterns, bind functions, SPIN RDF
Adding structure (class & property definitions & restrictions)
N/A - partially addressed in
1) generating resources from literals, and
2) interlinking data using GeoNames
class and property hierarchy, inverse properties
Designing sample queries, results visualization & data exploration
Grouping, contexts, Using literals to generate resources Multiple filters, grouping
Interlinking datasets / Using LOD resources
Using GeoNames country data as a third subgraph
to act as mediator in linking the two other subgraphs.
Union graph patterns, owl:sameAs, rdfs:label
N/A - 1 dataset
Designing complex queries
Using data from all subgraphs, calculating fractions, Subqueries, calculating fractions,...
  • Using OntoRefine to clean up non-RDF data
  • Transforming the data and loading it into a repository
  • Adding structure (class & property definitions & restrictions)
  • Designing sample queries, results visualization & data exploration
  • Interlinking datasets / Using LOD resources
  • Designing complex queries

Consultancy session

Attendees are asked to submit a form with their availability and goals/ideas about the consultancy session. These and their profile information is used to assign them to appropriate people to address their needs

Materials

All materials can be found at "\\ontonas\training\Designing Semantic Technology PoC"

Pre-class materials can be found at:

  • "\\ontonas\all-onto\training\Designing Semantic Technology PoC\Pre-class videos materials" for videos, lecture slides and video narrative
  • "\\ontonas\all-onto\training\Designing Semantic Technology PoC\SPARQL Exercises" for SPARQL exercises, dataset and sample solutions

Live session materials can be found at "\\ontonas\all-onto\training\Designing Semantic Technology PoC\Live session materials". Those include session narrative, datasets used, ontorefine commands used, SPARQL queries ran. A README.txt describes the necessary files and input parameters for each repository / workflow example.

Recordings of the two live sessions held on 2016/12/13 can be found at "\\ontonas\all-onto\training\Designing Semantic Technology PoC\Live session materials\20161213 sessions recordings".

Labels:
None
Enter labels to add to this page:
Please wait 
Looking for a label? Just start typing.