[OKFN-EN] understanding user difficulties with Description Logics

Paul Warren paul.w.warren at btinternet.com
Tue Mar 31 15:14:09 UTC 2015


Description Logics, such as the varieties of the Web Ontology Language
(OWL), are now the main languages used in the creation of ontologies.  Yet
they are recognised as posing difficulties for users.  I am writing for your
help with a study which seeks to identify those difficulties and to propose
ways of reducing them.  I have recently undertaken face-to-face experiments
with a number of people and am extending the study for wider participation.

 

There are four separate sections to the study, dealing with different
aspects of Description Logics.  Each section has been created as a separate
'quiz' consisting of eight questions.  Each question consists of a set of
axioms and a proposed inference.  Participants are required to indicate
whether the inference is valid or non-valid.  The questions use a simplified
form of Manchester OWL Syntax.  All you need to know about the OWL
constructs and the syntax used is contained in a brief handout, available
here
<http://people.kmi.open.ac.uk/warren/kmi-data/uploads/2015/03/handout_v3.pdf
>  (there is also a link in each of the quizzes).  Please have this handout
available as you answer the questions.

 

Here are the links to the four quizzes, covering four different themes:

.         object properties
<http://www.surveygizmo.com/s3/2039435/OWL-Quiz-Properties> 

.         negation and quantification
<http://www.surveygizmo.com/s3/2047772/OWL-quiz-negation-some-only> 

.         multiple quantification, i.e. the use of two quantifiers
<http://www.surveygizmo.com/s3/2056885/OWL-quiz-multiple-quantifiers> 

.         propositional logic, specifically
<http://www.surveygizmo.com/s3/2044995/OWL-quiz-not-and-or> not, and and or

 

Your help with this study would be greatly appreciated.  Each quiz should
take between 10 and 15 minutes.  Obviously, it would be great if you could
do all four.  However, if you can only do one or two, that would also be a
great help.  You can, of course, do the four quizzes over a period of time.
The deadline for the study is Friday April 24th.  Everyone who has taken
part so far has found the experience interesting.  All responses will be
treated confidentially and only aggregated results will be published.  The
intention is to compare the difficulty of the various questions, and hence
the difficulty of the underlying constructs.

 

Many thanks

 

Paul Warren

Knowledge Media Institute, Open University, U.K.

http://kmi.open.ac.uk/people/member/paul-warren 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/okfn-en/attachments/20150331/ee82e4dd/attachment-0003.html>


More information about the okfn-en mailing list