Semantic Web
WS 2022/23
Prof. Dr. Wolfgang May,
Lars Runge, M.Sc.
- Date and Time: Tue 14-16, Wed 10-12
- Room:
IFI 2.101 (North Campus)
- This year, DBIS will use mainly non-live teaching by pre-recordings.
There will be some live online meetings with BigBlueButton provided by GWDG;
the rooms/meetings can be entered via StudIP.
- Materials for self-studying (in english) will be linked below weekwise:
- Video recordings taken from winter term 2020/21 (as the "original" dates in the filenames indicate).
As long as the revised recordings are not yet available, the originals are available via StudIP:
Register for the "Semantic Web" in Winter Term 2020/21,
then you find them under "Meetings".
- PDF slides
- Please also read the general and technical information
about DBIS virtual teaching.
Lecture and Exercises mixed (see announcements on this page).
There will be non-mandatory exercise sheets whose solutions will be discussed as parts
of the lecture.
All materials and announcements can be found/linked HERE on the "blue DBIS pages".
Technical Data: Module Modul CS.M.Inf.1142,
4 SWS, 6 ECTS credits (Studies in Applied Informatics).
The module's home is the MSc studies in Applied CS.
It can also be credited in the BSc studies in Applied CS
(as "Vertiefung Datenbanken"),
and in several other studies:
BSc/MSc Wirtschaftsinformatik, Mathematics (BSc/MSc), Digital Humanities,
Teaching/2-Fach-Bachelor, PhD GAUSS, ...
Prerequisites
- Basic knowledge of databases (conceptual modeling, relational model, SQL),
background notions of "data model" in general, the idea of declarative
set-oriented query languages.
-
Knowledge in First-Order Logic as taught in "Formale Systeme" is recommended;
at least you should have some idea of it, and not be scared of formalisms.
Semantic Web is an example of applied First-Order Logic and model theory.
There is also the "sibling" lecture Deductive Databases
that is also applied First-Order Logic, but with a slightly different model theory.
-
XML: RDF/XML uses XML as representation, but requires only a little bit of
knowledge about XML. A short introduction to XML from that point of view will be given
in the lecture.
XML with DTD, XPath, XQuery, XSLT and XML Schema is the topic of the lecture
Semistructured Data and XML
(prospectively taking place again in Summer Term 2023).
Course Description
The course starts with the requirements on Web-wide data management
and querying in the early 2000s: Web-wide data, and intelligent data
integration aspects. In this context, the notions of metadata and
ontologies are discussed. The central topic of the course is then the
RDF data model, a graph data model with the corresponding
SPARQL query language. On the path towards knowledge bases, RDFS (RDF
Schema) and straightforward ontology portions of OWL are
introduced. On the technological side, Linked Open Data (LOD) is
presented. Finally, OWL-DL knowledge bases and reasoning that can
solve problems that are far beyond "simple" data management are
investigated. As practice-of-theory, this part of the course
illustrates the problems of open world negation (und universal
quantification).
- Short Review: Basic Notions of First-Order Logic
- RDF: N3 and RDF/XML format, semantics
- SPARQL: the query language for RDF data
- Linked Open Data (LOD): Web of Data and distributed querying.
- The Mondial database is
used as an example for RDF data.
Mondial LOD entry point.
- RDFS, OWL: having RDF data with additional reasoning
- Description Logics: the logic underlying OWL
- Practical experiments with RDF, Jena, Reasoners etc.
- An experimental Web interface can be found for RDF+OWL and SPARQL here
Dates & Topics
- First Meeting: Tue 25.10.
14-16, live online in BBB via StudIP:
Administrativa, Overview.
- Results of the polls wrt. preferred teaching format
-
Material for self-studying:
-
An overview of datamodel concepts and buzzwords
(data models XML and RDF, and mechanisms for "intelligent databases")
related to the DBIS lectures.
Database concepts and buzzwords recording
- Reasoning Motivation: the Einstein/Fish Puzzle
will be solved declaratively (totally different than in the "Deductive Databases" lecture).
Solve it by human reasoning, keep record of your solution steps. If you attended "Formal Systems":
what reasoning calculus would fit for it (and for sudokus, which are of a similar type)?
The following is a "virtual" schedule for structuring self-studying:
As long as the revised recordings are not yet available, the original
recordings from WS2020/21 are available via StudIP:
Register for the "Semantic Web" in Winter Term 2020/21,
then you find them under "Meetings".
The way towards the Semantic Web: Earlier Web Data Architectures, Formal Ontologies
- 26.10. (Recording from 4.11.2020): Web Architectures
- 1.11. (Recording from 10.11.2021): Ontologies
accompanying notes (this document will be incremental,
collecting all notes from the lecture taken with draw.io)
- 2.11. (Recording from 11.11.): Ontologies (cont'd)
(FOL formalism for ontologies here as an "applied formal system")
accompanying notes
- 8.11. (Recording from 17.11.2020): Ontologies (cont'd)
(FOL formalism for ontologies here as an "applied formal system")
accompanying notes
- 9.11. (Recording from 18.11.2020): Ontologies (cont'd)
notes from 17+18.11.
-
15.11. 14:15 (Optional) online live meeting. Questions and answers.
- 15.11. (Recording from 24.11.2020): Review of first-order logic and basic notions of
model theory and reasoning, Inference Systems
Note: on the recording, I started in german ... for 1:40 minutes,
then in English.
Slides: Introduction to Logics.
-
"Inofficial" Exercise Sheet 0 (Tableaux)
These (simple) exercises just review the FOL tableau calculus. Due to reasons of time,
they will not be discussed at length in the lecture. Here is the
[solution].
- 16.11. (Recording from 25.11.2020): Tableau calculus, Reasoning, Inference Systems
Further slides (from the Deductive Databases lecture) about
first-order logic and the relational calculus and
about reasoning and the FOL tableau calculus.
Data Model and Query Language for Data on the Web: RDF, RDFS and SPARQL
-
22.11. 14:15 (Optional) online live meeting. Questions? Outlook ...
- 22.11.2021 (Recording from 1.12.2020): RDF
(The BBB session/recording on 1.12.2020 BBB broke down after 6 minutes.
The lecture has afterwards been continued in the same room.
Thus, there are two separate recordings (timestamps 14:03, the one listed second,
and 14:25, the one listed first, with a duration of 6 and 80 minutes, respectively.)
Slides: RDF
Note: all example files are accessible in the RDF subdirectory of
this Web page
[ Download RDF.zip ]
[ SPARQL Web Service interface ]
- 23.11. (Recording from 2.12.2020): RDF, SPARQL
Exercise Sheet 1: SPARQL
-
29.11. 14:15 (Optional) online live meeting. Questions?
- 29.11. (Recording from 8.12.2020): SPARQL
Solution to Exercise Sheet 1
- 30.11. (Recording from 9.12.2020): SPARQL 1.1, SPARQL Formal Semantics
-
6.12. 14:15 (Optional) online live meeting. Questions?
- 6.12. (Recording from 15.12.2020): SPARQL Formal Semantics
Exercise Sheet 2: SPARQL Formal Semantics
- 7.12. (Recording from 16.12.2020): RDF: Blank Nodes, Tree vs. Graph Data Models & Logic,
RDFS
Slides: RDFS
-
13.12. 14:15 (Optional) online live meeting. Questions?
- 13.12. (Recording from 22.12.2020): RDFS Model Theory
- 14.12.2022 (Recording from 23.12.2020): RDFS Reasoning, querying etc (this chapter until the end)
-
Solution of Exercise Sheet 2 (Recording from 3.1.2021)
Data on the Web/Linked Open Data
This section gives an overview of practical usage issues
- 20.12. (Recording from 4.1.2021):
RDF/XML: the "back link" to the XML world which was the driving force
to create the "Semantic Web": give meaning to XML element names and attribute
names.
RDF/XML ist nothing conceptually new,
but (if one knows XML well) mainly craft, like using URIs, and
expansion of element names to URIs with namespaces, and xml:base.
Its design illustrates the compromise between (existing) IT technology
(concrete XML as tree-structured document-data, URI/URI expansion), and the
abstract data model of an RDF graph, and the (common to both worlds) aspect
of namespaces. The result is a lot of syntax that finally results in a feasible IT concept.
Slides: RDF/XML
- 21.12.: Playing section: Linked Open Data (Recording from 7.1.2021).
Data on the Web, Web/HTTP Technology, accessing it as data triples and with SPARQL.
Slides: LOD
The experimental Mondial LOD Service can be found at
http://www.semwebtech.org/mondial/10
- 21.12.: Reification and Modeling/Wikidata (Recording from 9.1.2021).
(Slides "LOD" from above)
Wikidata is an example of a Knowledge Graph for LOD data
that uses flat RDF, but not RDFS and OWL since that would constrain
the expressiveness to much (basically, the problem that penguin
is at the same time (i) a class that contains individuals, and (ii) an individual
of the class Species that has user-defined properties of its own).
HTML Germany page in Wikidata
Wikidata SPARQL query interface
Semantic Web with Reasoning: Description Logics and OWL
-
10.1.2023, 14:15 (Optional) online live meeting. Questions? Outlook ...
(Nobody attended, so there was no recording.)
- 10./11.1.2023 (Recordings from 12./13.1.2021): Description Logics
Slides: DL
- 17./18.1. (Recording from 19./20.1.2021): OWL
Slides: OWL
-
24.1.2023, 14:15 (Optional) online live meeting. Questions? Outlook ...
- 24./25.1. (Recordings from 26./27.1.2021): OWL ...
Exercise Sheet 3: OWL
- 31.1./1.2. (Recordings from 2./3.2.2021): OWL, OWL 2.0
Slides: OWL 2.0
- ... and now you can also try to solve the Einstein/Fish Puzzle.
- 7.2. (Recording from 9.2.2021): OWL 2.0 (cont'd)
- 8.2. (Recording from 10.2.2021): Discussion of Exercise Sheet 3, OWL 2.0
Solutions to Exercise Sheet 3 (OWL) [solution]
- ... finally, there is also the solution to the Fishpuzzle in OWL.
There are two modelings. At the beginning (more than 10 years ago, a very old version of pellet)
the short one took about 5 minutes, and the long one took 40 hours. Today, both run in
some seconds, which demonstrates mainly the progress in the reasoners' internal strategy:
- fishpuzzleLong.n3 is a very detailed
and intuitive specification:
Part 1: Specification of a row of 5 houses.
Part 2: Specification of the properties.
Part 3: Specification of the constraints.
- fishpuzzleLong.sparql is the corresponding query
- fishpuzzleShort.n3 is a shorter encoding:
Part 1: Specification of a row of 5 houses as above.
Part 2: tricky encoding of the properties. Instead of assigning to each house a color, a person,
a brand of cigarettes, a drink, and a pet via explicit RDF edges,
they are declared to be sets that are identified/mapped to each other.
Part 3: specification of the constraints. Simpler as before, since only the equivalence
classes have to be considered.
The reasoner then has just to compute the matching (actually, this task is very similar
to what StableModels does in the same situation based on clauses/disjunctions).
- fishpuzzleShort.sparql is the corresponding query.
- End of lecture period: 10.2.2023
The complete slide set can be found here
(including the Semantic Web Lab Course Slides). The slides of the
SSD&XML lecture can also be found there. Knowledge of XML is only
required so far as RDF/XML is (in addition to the N3 format) a
possible representation of RDF data. One should be able to
"understand" an XML document. XPath/XQuery and XSLT are not required.
Exams
- Oral exams, (nearly) any timepoint (see below)
- Exam procedure: about 30-40 minutes, most probably online . Candidates start
with talking about a topic of their choice from the lecture (5-10 minutes),
then questions+answers, including sketches on paper develops dynamically.
The 5-10 minutes talk at the beginning should give me as an examiner a good
impression of your knowledge, and a good starting point to assess your
knowledge with further questions (usually starting with the chosen topic, and then
also going to other topics from the lecture).
Languages in oral exams: English and/or German.
- With this year, we try a modified strategy for oral exams:
Oral exams, basically whenever needed:
- Registration via FlexNow:
-
The exam regulations
(Allgemeine
Prüfungsordnung BSc/MSc Göttingen (2013), Par.10b)
and the FlexNow system (that we must use) are not very appropriate
for administrating flexible individual oral exams. We solve this as follows:
- Oral exams and the formal registration in FlexNow are "always"
possible, this means, also during semesters where the lecture does not
take place.
- Deregistration in FlexNow is not possible. So register in FlexNow
only when you are sure that you want to do the exam. You must
be registered when actually doing the exam.
- The "FlexNow Exam" for winter term lectures is thus configured as
follows: Registration between Feb 1st and Jun 30th (De-registration
until Feb 1st, means "not possible"). If you want to do the exam later,
use the subsequent analogous exam slot (always Jul 1st - Jan 31th).
Note that both slots are listed under "Winter Term 2021/22"
in FlexNow because they refer administratively to the winter term lecture.
- For your actual individual exam appointment, contact
may at informatik.uni-goettingen.de for a concrete appointment,
usually 3-4 weeks before the exam, specifying which week and maybe even
what day you prefer and morning/afternoon.
(We will become aware of your registration only via this mail - FlexNow
does not notify us about incoming registrations - and usually, we do not
look inside it actively.)
Background Literature
P. Hitzler, M. Krötzsch, S. Rudolph, Y. Sure: "Semantic Web - Grundlagen" (in German).
Springer eXamen.press, 2008;
ISBN 978-3-540-33994-6.
The (german language) book covers nearly exactly the contents of the lecture and also
contains an introduction to first-order logic in the appendix.
P. Hitzler, M. Krötzsch, S. Rudolph, Y. Sure: "Foundations of Semantic Web Technologies"
(in English).
Chapman & Hall/CRC, 2009;
ISBN: 9781420090505
The (english language) book covers nearly exactly the contents of the lecture and also
contains an introduction to first-order logic in the appendix.
For the part on (first order) logic, and textbook on foundations of logic from the library
(e.g. "Logik für Informatiker" (in German) von Uwe Schöning) or the manuscript
"Formale Systeme" by
Peter H. Schmitt (Uni Karlsruhe) (Kap. 1-5) can be used.
Some Links
If you experience any problems (forgotten chmod, wrong paths, forgotten updates etc.),
please notify us.
Web-wide Services
Professional Tools
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 for querying:
- Download most recent version
(incl. OWL2, partially SPARQL 1.1, based on Jena 2.10 and compatible Openllet,
migrated to log4j-2.16), Dec. 2021.
- in the CIP Pool located at
/afs/informatik.uni-goettingen.de/course/semweb-lecture/JENA-API/semweb.jar
- Experimental Web interface
- Openllet Homepage
(Open Source tool based on the last freely-available version of Pellet)
- Pellet itself has been turned into a commercial product at Stardog.
LOD: Accessing RDF Data in the Web
- rapper: a tool that accesses a Web page in RDF-reading mode to get RDF triples: e.g.
rapper http://sws.geonames.org/3017382/
The Mondial database in RDF format can be found at
http://www.dbis.informatik.uni-goettingen.de/Mondial/#RDF.
Call e.g.
jena -q -qf mondial-query.sparql
or
jena -pellet -q -qf mondial-meta-query.sparql
Usage in the CIP Pool
From the CIP Pool computers at the IFI (ground floor or log in from remote), the software and resources are directly
accessible:
- log in from remote to login.stud.informatik.uni-goettingen.de (Linux: ssh, Windows: puTTY)
- log through to one of the individual computers (e.g. ssh c032)
- set the alias in your .bashrc file:
alias jena='java -jar /afs/informatik.uni-goettingen.de/course/semweb-lecture/JENA-API/semweb.jar'
- The lecture's RDF directory with the n3 files can be found at
/afs/informatik.uni-goettingen.de/user/d/dbisuser/public_html/teaching/SemWeb/RDF
- The Mondial files can be found at
/afs/informatik.uni-goettingen.de/user/d/dbisuser/public_html/Mondial
/afs/informatik.uni-goettingen.de/user/d/dbisuser/public_html/Mondial/Mondial-RDF
|
