Note: this course does not give an introduction to Semantic Web, but builds upon the knowledge from the Semantic Web WS 2019/20 lecture. The next introductory course to Semantic Web will prospectively take place in WS 2020/21.

Prof. Dr. Wolfgang May
Lars Runge, M.Sc., Sebastian Schrage, M.Sc.

• Date and Time: Friday 14-18, depending on how long the course takes. The course meeting will not take place every week.
  
• Virtual Meetings: We will use BigBlueButton provided by GWDG; the rooms/meetings can be entered via StudIP.
  Please also read the general and technical information about DBIS virtual teaching.

The course yields 6 ECTS-credits; it is graded (=benotet)

Prerequisites:

• Successful participation in the module "Semantic Web"
• Successful participation in the "General Programming Lab/Allgemeines Programmierpraktikum" (or an equivalent course).

Course Description

The practical course builds upon the lecture Semantic Web. Among other sources, the Mondial database in RDF format is used.

Documentation: use the slides from the Semantic Web lecture (the full slide set can be found here)

The course takes place in groups of 3-4 persons. There will be several units.
Usually, for every unit there is a course meeting, and additionally individual meetings with the groups.

Topics: Warm-Up: SPARQL revisited, Linked Open Data, distributed queries in SPARQL (e.g. using Wikidata), RDF/OWL with Java/Jena, OWL and hybrid OWL+ (Datalog-style) rules reasoning, Tools: OWLAPI (Manchester Syntax) +Protege, RDF4J (ex-Sesame), Querying and analyzing OWL Metadata (which is not easily possible with only SPARQL), (maybe: Web Services, if not taught before in the XML lab course)

Time Schedule

• (Friday 17.4.: No meeting due to the 1 week delay of the semester begin)
• 1. Meeting 24.4. (Friday), 14-16 ct SR 2.101: Introduction, Formalities, SPARQL revisited
  • Slides (from the lecture): RDF, SPARQL, RDFS
  • Exercise Sheet 1: SPARQL - to be discussed until 8.5.2020.
• 1.5.: No course because of Labor Day
• 8.5.: Discussion of the first exercise sheet
  
• 15.5.: Linked Open Data. Mondial as LOD example.
  
  • Slides (from the lecture): Linked Open Data (LOD)
  • Other examples: Wikidata and its Modeling. DBPedia, Yago, ...
    Some examples for HTML LOD SPARQL endpoints with Web forms to explore these sources:
    • Mondial
    • Wikidata
    • Insee.fr
    • geonames (http://www.lotico.com:3030/lotico/sparql) does not have an HTML interface
    • DBpedia
    • Yago
    Syntax of SPARQL functions and operators:
    • W3C SPARQL Recommendation
    • Wikibooks, with wikidata examples
    An URL-encoding service can be found here
  • Exercise Sheet 2: Linked Open Data to be discussed until 29.5.
• 22.5.: RDF/RDFS/OWL and SPARQL and Java. The Jena API.
  • Slides: Jena API
  • Example code: Jena Basics.java
  • Exercise Sheet 3: RDF/RDFS/OWL and SPARQL with Java/Jena to be discussed until 12.6.
    Country data for Catalonia: catdata.rdf
• 29.5.: Discussion of exercise sheet 2
• 5.6.: Rule-based Reasoning: RDFS, OWL-RL, Rules in Jena; the internal rule-based Jena reasoner.
  • Slides: OWA vs. CWA, SWRL, Rule-based Reasoning in Jena
  • Background Slides (material from the lecture): OWL
  • Exercise Sheet 4: Reasoners, Rules and Ontology Updates to be discussed until 26.6
    
• 12.6 Discussion of exercise sheet 3
• 25.6.: OWL-API (Manchester-Syntax for OWL/DL ), Hermit, Protege (Ontology editing; supports also SWRL rules in Manchester Syntax)
  • Slides: OwlAPI, Hermit, Protege
  • Exercise Sheet 5: OWLAPI, Protege
• 17.7., 14:00: Discussion of Exercise Sheet 4
• TO BE EXTENDED

Software

Jena: RDF and SPARQL

Apache JENA ( https://jena.apache.org/ ) is a free and open source Java framework for building Semantic Web and Linked Data applications.
The course uses a lightweight housemade shell interface to Jena with Pellet for querying in a still experimental unstable version:
If something does not work, send us an email.

• Download most recent version based on Jena 3.10 and Openllet (the open source continuation of Pellet-2)
• in the CIP Pool located at /afs/informatik.uni-goettingen.de/course/semweb-lecture/JENA-API/semweb-23-mai-2019.jar
• SPARQL Web interface (via the Mondial-LOD service)
  (it adds rdf/rdfs/owl/mondial prefixes automatically)

• requires Java 8
• set alias (bashrc etc.)

   alias jena='java -jar /afs/informatik.uni-goettingen.de/course/semweb-lecture/JENA-API/semweb-23-mai-2019.jar'

• query: (if=input-files, qf=query-file, e.g. in SPARQL)
  jena -q -if inputfiles -qf queryfile
• general options:
  -il: input language (not necessary, is auto-detected)
  -if: input files
• query options:
  -q: query
  -il, -if: as above
  -qf: query-file
  -ofo optimize-federated-off (concerns the eval. of SERVICE clauses in SPARQL)
• transform options:
  -t: transform
  -ol: output format (allows RDF/XML RDF/XML-ABBREV N-TRIPLE N3-PLAIN N3-PP N3-TRIPLE N3 TURTLE; N3 is default)
• export class tree; options:
  -e: export class tree (gives some insight for debugging an ontology ...)
  -il, -if: as above
• reasoner options (for -q and -e):
  activate reasoning; default: internal rule-based reasoner: option -inf (for "inference")
  or use the pellet/openllet class that comes with the semweb.jar:

   jena -q -inf -qf query-filename
   jena -q -pellet -qf query-filename

Raptor RDF Syntax Library - Raptor RDF parser utility

• access URLs as RDF client rapper instructions