StudyGuideSWE Study Guide

Study Guide

Please use the following guideline for the preparation for the exam.

Details on the Exam

1. The available time for the exam will be 120 minutes.

2. There will be three questions out of which two must be chosen. Do not answer more than two questions.

3. The exam will be given in German and can be answered in German or English at your choice.

Topics and Contents

1. You will be asked to answer questions on technical or economical aspects of the Semantic Web vision and technology stack, and on the interplay between those two.

2. Relevant for the exam are the slides and contents of all units.

3. The exact syntax of e.g. W3C languages or particular ontologies is not expected. However, you need to have understood the various technology components to a degree that allows an informed discussion of the potential and problems in the business usage of the Semantic Web.

Criteria

The following criteria will be used for grading: Total points per question: 15

a) Writing, Structure, Flow of Argument (0...5 Points)

b) Depth and Quality of the Judgment (0...5 Points)

c) Technical correctness (0...5 Points)

Important: Make sure you are precise and detailed in your answers. No bla-bla or sales talk, please!  Be clear and concise. Tables often help getting to the point. Use a meaningful structure for your text (i.e., use section headers and subsections, if appropriate).

Important Topics and Aspects of Particular Importance

Lecture 1: Introduction

Explain the technical limitations of the current Web for search and other operations on a global basis. Which exact problems are the causes of unsatisfying results?

There exists a "shallow" Web (i.e., pages) and a "deep" Web (i.e., database content available via the Web). Discuss whether, and if so, how, we must address this difference when implementing the Semantic Web vision.

There are two ways of improving the ability of computer systems to process tasks: We can make the problems more suitable for computer processing, or we can increase the ability of computers to process unstructured input data (e.g. natural language). Which of these two options is the main direction of the Semantic Web movement?

Explain the three key components of the mainstream Semantic Web approach, and detail what each component will actually contribute.

Name and explain six effects that ontologies may have on exchanging and processing information on the Web.

Lecture 2: Resources and Their Identification on the World Wide Web

Explain the differences between URI, URL, an URN.

What are "cool" URIs, and why are they important? Give examples of good and bad URIs in that sense.

What is the difference between information resources and non-information resources?

What are network effects (externalities)? Are there positive network effects with regard to URIs?

Lecture 3: From Knowledge Organization Systems to Ontologies

What is an ontology in computer science?

Explain how popular classification hierarchies are different from true ontologies.

Why are Knowledge Organization Systems (KOS) a promising resource for building ontologies for the Semantic Web?

What are popular structural limitations (e.g. anomalies) of KOS?

Lecture 4: Ontological Engineering, OntoNaviERP, WordNet

Explain the basic steps of ontology engineering.

What are competency questions, and what do they help define?

Why is ontology engineering unfeasible when the relevant domain uses a large and detailed terminology?

What is WordNet, and how can it help improve the construction of ontologies?

What is the difference between a concept and a word in a natural language?

Which approaches may help reduce the effort for building large, lightweight ontologies (think of the OntoNaviERP approach)?

Lecture 5: Ontology Economics

Ontologies are inherently community activities, i.e., many individuals must contribute time and other resources. Explain the role of proper incentives for the construction and broad usage of ontologies. If possible, explain the problem from the perspective of a single individual.

Discuss the role of positive network externalities with regard to ontologies on the Semantic Web. How could that explain the current dominance of small, rather lightweight ontologies?

(This whole unit is a very important topic for the exam.)

‎Lecture 6: The GoodRelations Ontology: Semantic Web-based E-Commerce

What is the purpose of the GoodRelations ontology?

Explain the most important conceptual elements of the GoodRelations ontology.

Explain the immediate benefit for any Web shop in the World to use the GoodRelations ontology.

(see http://www.heppnetz.de/projects/goodrelations/webcast/ for a 15-minute video on the topic).

‎Lecture 7: Accessing and Managing RDF Data: Storage, Reasoning, and Query Languages

Explain the role of a repository and a reasoner in managing RDF data.

What does a reasoner contribute to using Semantic Web data?

Why can't we simply use XML query languages on RDF/XML data? (Hint: Keep in mind that the same facts can be represented in many different ways in RDF/XML)

What is SPARQL?

Lecture 8: The Web Ontology Language OWL

What is the relationship between RDF and RDFS, and between RDFS and OWL?

There are different language variants of OWL: Lite, DL, and Full. Which one would you choose for which type of application?

‎Lecture 9: Semantic Web Services

Why is it important to include Web Services, i.e., computational functionality that can be invoked via the Web, in the Semantic Web vision?

Explain the major difference between annotating Web contents (data) vs. Web functionality (behavior).

Explain how ontologies can, in theory, support services discovery and service composition.

The early Semantic Web Services community has stressed that we must annotate all aspects of services. However, this will pay out only if the total benefits from discovery or composition outweigh the total effort for creating and maintaining the annotations. Quite obviously, the actual frequency of searching for a new service  will determine whether Semantic Web services will make sense. Discuss that aspect.

‎Lecture 10: RDFa and GoodRelations

What is RDFa and how is it related to RDF and to RDF/XML?

Why is RDFa important for the broad adoption of the Semantic Web by average Web masters? (=> embedding Semantic Web data into existing XHTML content!)

How can RDFa help to keep Semantic Web data (for machines) and Web data (for humans) in alignment, in particular for literal values?