The Semantic Web is an extension of the World Wide Web in which information is represented in a machine-understandable way. In addition, hyperlinks of the Semantic Web connect semantically related information based on ontologies. Thanks to these properties, it becomes possible to automate reasoning over a set of hyerlinked information on the Semantic Web using computational agents. In this paper, we propose to use a Semantic Web reasoner in order to help students understand concepts in informatics. More speci cally, our approach consists of three parts. First, learning materials are collected from DBpedia (https://wiki.dbpedia.org) that is an online knowledge base whose contents are provided in several formats including RDF, Notation3, etc. Second, a teacher de nes inference rules according to which new facts can be derived from the set of learning materials taken from DBpedia. Third, a learner executes a system called EBRS that is a simple-to-use interface for a well-known Semantic Web reasoner, EYE (http://eulersharp.sourceforge.net/). EBRS shows the result of reasoning on a screen so that learners can understand the relationship of learning materials easily. The theme of this research is to exploit Semantic Web technologies such as RDF, ontologies, reasoning programs, etc in order to support informatics education.
Ever since the Semantic Web [
F. Author et al.
The basic idea behind our approach is to use a reasoning system over the set
of educational linked data on the Semantic Web so that the system can show a list
of inter-related concpets which is the reasoning result. Since the data on the
Semantic Web are represented in a machine-understandable way, the identi cation
of two or more related concepts can be automated using inference rules. To this
end, we use a system called EBRS [
More speci cally, Algorithm.n3 is downloaded from DBpedia and it contains a list of triples in which each triple takes on the form of \subject-predicateobject". In a triple, a predicate can be interpreated as a hyperlink that connects a subject and an object. A part of Algorithm.n3 where the predicate \hypernim" is used is shown below. \ns0" in front of \hypernim" denotes the namespace http://purl.org/linguistics/gold/ in which this word is de ned. For example, the rst line means that Multiple signal classi cation and Algorithm are in a hypernim relation. ... dbr:Multiple_signal_classification ns0:hypernym dbr:Algorithm . dbr:Intelligent_Water_Drops_algorithm ns0:hypernym dbr:Algorithm . dbr:Bareiss_algorithm ns0:hypernym dbr:Algorithm . dbr:Weak_NP-completeness ns0:hypernym dbr:Algorithm . ...
Algorithmrule.n3 is shown below and it is the le that contains inference rules that a teacher can de ne. This le consists of three namespaces followed by an inference rule that states that if \s" and Algorithm are in a hypernim relation, then \s" and Complexity are related. There can be various possible inference rules that can be de ned here and which inference rules can be de ned depends on the structure of learning materials. It is this part that a teacher should be careful about in order to provide students with helpful knowledge that can be derived by reasoning.
Application of Semantic Web technologies in Informatics Education {?s ns0:hypernym dbr:Algorithm.}=>{?s "is_related" dbr:Complexity.}
EBRS shows the result of applying the inference rule in Algorithmrule.n3 to Algorithm.n3 on a screen and a part of the result is shown below. Note that the following four triples are newly generated based on the reasoning process. This can help learner understand concepts and their relationship because it is explicitly stated that they are related. In other words, before the reasoning process, these exist separately without being connected in Algorithm.n3 le, but the reasoner identi es them and provides them with leaners as an explicit list of triples so that learners can understand the interrelationship of the learning materials easily. ... dbr:Multiple_signal_classification "is_related" dbr:Complexity. dbr:Intelligent_Water_Drops_algorithm "is_related" dbr:Complexity. dbr:Bareiss_algorithm "is_related" dbr:Complexity. dbr:Weak_NP-completeness "is_related" dbr:Complexity. ...
Conclusions
In this paper, we report our ongoing project which aims at exploiting Semantic
Web technologies in informatics education. The proposed approach is to use a
machine-understandable format such as Notation3 to represent learning
materials and present students with reasoning results over the set of learning materials
so that implicit hyperlinks that exist among the set of learning materials can
become explicit. This can help students understand the relationship among a
set of concepts that they are not familiar with. Currently, we are investigating
ways by which concepts in theoretical computer science [