![On the Use of Linked Open Data in Education:
Current and Future Practices
Mathieu d’Aquin(B)
Knowledge Media Institute, The Open University,
Walton Hall, Milton Keynes, UK
mathieu.daquin@open.ac.uk
Abstract. Education has often been a keen adopter of new informa-
tion and communication technologies. This is not surprising given that
education is all about informing and communicating. Traditionally, edu-
cational institutions produce large volumes of data, much of which is
publicly available, either because it is useful to communicate (e.g. the
course catalogue) or because of external policies (e.g. reports to funding
bodies). Considering the distribution and variety of providers (universi-
ties, schools, governments), topics (disciplines and types of educational
data) and users (students, teachers, parents), education therefore repre-
sents a perfect use case for Linked Open Data. In this chapter, we look
at the growing practices in using Linked Open Data in education, and
how this trend is opening up opportunities for new services and new
scenarios.
Keywords: Linked data · Semantic web · Education · Learning
1 Why Using Linked Data in Education
Traditionally, educational institutions produce large volumes of data, much of
which is publicly available, either because it is useful to communicate (e.g.,
the course catalogue) or because of external policies (e.g., reports to funding
bodies). In this context, open data has an important role to play. Implementing
open data through Linked Data technologies can be summarized as using the
web both as a channel to access data (through URIs supporting the delivery of
structured information) and as a platform for the representation and integration
of data (through creating a graph of links between these data URIs). Considering
the distribution and variety of providers (universities, schools, governments),
topics (disciplines and types of educational data) and users (students, teachers,
parents), education also represents a perfect use case for Linked Open Data [7].
Indeed, the basic idea of Linked Data [9] is to use the architecture of the Web
to share, distribute and interconnect data from various origins into a common,
online environment. It is based on the basic principle that raw data objects are
identified and accessible, similarly to webpages, through Web addresses (URIs),
that deliver the information in a structured, processable and linkable way.
c
Springer International Publishing Switzerland 2016
D. Mouromtsev and M. d’Aquin (Eds.): Open Data for Education, LNCS 9500, pp. 3–15, 2016.
DOI: 10.1007/978-3-319-30493-9 1](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-13-320.jpg)
![4 M. d’Aquin
This approach has been very successful in the last few years, especially as a
base method for the publication of open data on the Web. Linked Data has
been adopted by government agencies in several countries (prominently, in the
UK and the US) for transparency and public information purposes, by cultural
heritage institutions such as libraries and museum to provide more processable
and integrated information about their collections (see the Europeana project1
for example, or the British Museum Collection2
), by companies in publishing
(for example at Nature, or Elsevier), broadcasting (for example at the BBC),
or retail (for example at BestBuy). As we will see later in this chapter, there is
a growing trend in the use of Linked Data specifically for education, with uni-
versities in particular making their public information (academic programmes,
research outputs, facilities, etc.) available as linked data on the Web (see for
example LinkedUniversities.org).
2 Linked Data - In More Details
The foundation of the Web is that it is a network of documents connected by
hyperlinks. Each document is identified by a Web address, a URI, and might rep-
resent a document which content is encoded using a standard, universally read-
able format (most commonly HTML). The foundation of Linked Data is that data
objects on the Web are identified, similarly to documents, by URIs. The represen-
tation of the data – i.e. the information associated with a data object – is then
represented by Web links, which can themselves be characterised by URIs. This
makes it possible to represent information in such a way that it is materialised as
a graph, where nodes are URIs or literal data values (strings, numbers) and the
edges are links between them.
For example, a university like The Open University3
publishes information
about the courses it offers through its website, as well as using linked data [3]. It
achieves that through assigning to every course a dedicated URI that acts both
as an identifier for the course on the Web, and as a way to address information
about this course. For example, http://data.open.ac.uk/course/aa100 is the URI
for the course with code AA100, which is an undergraduate (level 1) course in
Arts and Humanities, entitled “The arts past and present”. Through the links
between this URI and others, information about this course is being represented
regarding the topics and description of the course, where it is available, how it is
assessed, what course material and open educational resources relate to it, etc.
(see Fig. 1).
While most of the other data objects it relates to are also identified by URIs
within the domain of the Open University, it is important to remark here that
it links to other data sources, such as the UK government’s information about
The Open University or information provided by the Geonames platform about
1
http://www.europeana.eu/.
2
http://collection.britishmuseum.org/.
3
http://www.open.ac.uk.](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-14-320.jpg)
![On the Use of Linked Open Data in Education 5
Fig. 1. Extract of the Linked Data (RDF) representation of the course AA100 “The
Arts Past and Present” at The Open University (from data.open.ac.uk).
the countries in which the course is available. This demonstrates how, from
these basic principles, information originating from widely different systems and
sources can be seamlessly integrated.
Following the base principles described above, the most basic technology
employed to implement linked data is a web-enabled, graph-based data rep-
resentation language: RDF (Resource Description Framework [10]). RDF is in
principle related to XML, but dedicated to the representation of graphs where
nodes are URIs or literal values, and edges are links labelled by URIs. It has
different syntaxes, including an XML-based one, but also others based on listing
the triples [subject, predicate, objects] forming the links in the data.
Another important component of the technological stack for linked data is
the one of vocabularies. Indeed, it is important that data should be shareable
and reusable in a common way across sources and systems. To address that,
languages such as RDF-Schema [2] and OWL (the Web Ontology Language [11])
allow one to define the types of objects that can be encountered in the data (the
classes, e.g. Course, Person, Country, etc.), as well as the types of relationships
that connect these types of objects (the properties, e.g. location, title, employer,
author, etc.).
Finally, another important element of linked data is the way in which, still
relying uniquely on the basic mechanisms of the Web, the data can be consumed.
As we already mentioned, URIs on linked data can be requested to obtain RDF
(most often in its XML syntax). When more flexibility is required, many of the
existing linked data sources offer data endpoint using the standard querying
language and protocol for RDF/Linked Data: SPARQL [12]. Briefly, SPARQL
is both a query language made explicitly to fit the graph data model of RDF,
and a Web protocol dictating the way in which a SPARQL endpoint should be
accessed and queried on the Web. For example, the query:](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-15-320.jpg)
![On the Use of Linked Open Data in Education 7
3.1 The LinkedUp Project and the LinkedUp Catalogue
of Educational Datasets
The LinkedUp Project (Linking Web data for education [8]) was an EU FP7
Coordination and Support Action running from November 2012 to November
2014 which looked at issues around open data in education, with the aim of push-
ing forward the exploitation of the vast amounts of public, open data available
on the Web. The project comprised six pan-European consortium partners led
by the L3S Research Center of the Gottfried Wilhelm Leibniz Universitt Han-
nover and consisting of the Open University UK, the Open Knowledge Foun-
dation, Elsevier, the Open Universiteit Nederland and eXact learning LCMS.
The project also had a number of associated partners with an interest in the
project including the Commonwealth of Learning, Canada, and the Department
of Informatics, PUC-Rio, Brazil.
To aid awareness and use of open and linked data in education, the project
created and has continuously maintained a catalogue and repository of data
relevant and useful to education scenarios. The goal of the LinkedUp Dataset
Catalog (or Linked Education Cloud7
) is to collect and make available, ideally
in an easily usable form, all sorts of data sources of relevance to education, pro-
viding a shared, evolving resource for the community interested in Web data
for education (see Fig. 2). During the project, the technical team has enabled
and encouraged content- and data-providers to contribute new material to the
LinkedUp Dataset Catalog through a series of hands-on workshops and the pro-
motion of community documentation on LinkedUp tools, workflows and lessons
learned.
Datahub.io is probably the most popular registry of global catalogues of
datasets and forms the heart of the Linked Open Data cloud. In the interest
of integrating with other ongoing open data effort, rather than developing in
isolation, the LinkedUp Data Catalog has been created as part of Datahub.io.
It takes the form of a community group in which any dataset can be included,
provided that it is relevant, and the datasets in this group are also visible glob-
ally on the Datahub.io portal. Every dataset is described with a set of basic
metadata and assigned resources. This makes it possible to search for datasets
and employ faceted browsing of the results both globally or specifically in the
Linked Education Cloud. For example, one could search for the word ‘univer-
sity’ in the Linked Education Cloud, and obtain datasets that explicitly mention
‘university’ in their metadata. These results can be further reduced with filters,
for example to include only the ones that provide an example resource in the
RDF/XML format.
One of the key aspects of the design of the LinkedUp catalogue is that it itself
creates a Linked Data resource in addition to the use of Datahub.io. Indeed,
once datasets have been identified and registered, basic metadata related to each
of them, as well as information about their content, are automatically extracted
from Datahub.io and from their SPARQL endpoint. This information is then
7
http://data.linkededucation.org/linkedup/catalog/.](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-17-320.jpg)
![8 M. d’Aquin
Fig. 2. Screenshots of the Web interface to the LinkedUp Catalogue of Datasets for
Education for browsing datasets.
represented in RDF using the VoID vocabulary8
[1] and made available, in a
Linked Data way, through a SPARQL endpoint. It is this SPARQL endpoint
that we use and interrogate to analyse the characteristics of existing datasets for
education in the next section.
3.2 The State of Linked Data in Education
An initial analysis of an earlier version of the catalogue was shown in [5]. It
focused on the connection between datasets through their reuse of common
vocabulary elements. The core figures from that paper are reproduced in Fig. 3
below, showing the network of datasets and there partitioning through the com-
mon reuse of vocabulary elements, and the most commonly used classes/concepts
in these datasets, connected by their co-occurrence.
The current version of the LinkedUp catalogue is however much bigger: It
references 56 different datasets, each with their own SPARQL endpoint. Datasets
are obtained from a variety of sources. As can be seen from Fig. 4 however,
they essentially originate from either universities publishing their own data, or
from repositories of educational or research resources. Government open data
also contribute significantly to datasets related to education, with for example
statistics about the registration and results of educational institutions.
A simple aspect one might want to look at when analysing datasets about
education from the LinkedUp Catalogue is the variety of sizes that the datasets
represent. Each dataset might in particular be divided into multiple sub-graphs,
8
http://www.w3.org/TR/void/.](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-18-320.jpg)
![On the Use of Linked Open Data in Education 9
Fig. 3. Dataset (RDF graphs) connected by their reuse of common classes, and common
concepts (classes) connected by their co-occurrence (from [5]).
Fig. 4. Number of datasets from different areas in the LinkedUp Data Catalogue.
which might relate to different topics or originate from different sources. As
shown in Fig. 5 (which only includes datasets with more than one sub-graph)
the number of graphs included in each dataset can vary enormously (from only
1, to several thousands) depending on the way the dataset has been designed
and constructed. For example, the biggest one in number of graphs from Fig. 5
(SEEK-AT-WD) is constituted through crowdsourcing, and assigns a different
graph to each contribution. Some universities would include in one graph all the
information about all the courses they offer, while others might create a graph
for the representation of each course. As can be seen however, besides datasets
with very large numbers of graphs, or small datasets focusing on a small number
of topics, most datasets are structured using 10 to 100 graphs corresponding to
different aspects of the data (e.g. modules, resources, people, facilities, etc.).
For information, the chart in Fig. 5 is generated from the results of the fol-
lowing query on the SPARQL endpoint of the LinkedUp catalogue9
:
9
http://data.linkededucation.org/linkedup/catalog/sparql/.](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-19-320.jpg)
![On the Use of Linked Open Data in Education 13
Fig. 8. 40 most common properties amongst the datasets of the LinkedUp Data Cat-
alogue (in number of datasets).
select distinct ?c (count(distinct ?d) as ?n) where {
graph http://data.linkededucation.org/linkedup/catalogue/void {
?d a http://rdfs.org/ns/void#Dataset
{{?d http://rdfs.org/ns/void#subset ?sg.
?sg http://rdfs.org/ns/void#classPartition ?cp.
?cp http://rdfs.org/ns/void#class ?c}
UNION
{?d http://rdfs.org/ns/void#classPartition ?cp.
?cp http://rdfs.org/ns/void#class ?c} }
}
} group by ?c order by desc(?n) limit 20
4 Use and Future of Linked Data in Education
The analysis described above gives a general overview of the state of Linked Data
in education, of the way it is being used and how it can grow further. Indeed,
these datasets represent pioneering initiatives that should carry on growing, and
through understanding the expending practices of Linked Data in education
within these dataset, other can learn from them and have an easier entry point
to join the Linked Education CLoud. A key for this to happen however is for
these practices to be better shared. Indeed, another element shown in the analysis
above is that many of the initiatives at the origin of the considered datasets have
been developed in isolation from each other, with different modeling principles,
designs and vocabularies being used. While it is the nature of Linked Data (and
to an extent of the Web) that it should allow this kind of distribution, some
concertation is still required to ensure that the resulting datasets can be used
jointly in a way which is sufficiently cohesive (see [5]). Several initiatives have](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-23-320.jpg)
![14 M. d’Aquin
emerged that are trying to address this, among which LinkedUniversities.org,
LRMI14
and the W3C Open Linked Education community group15
.
An important aspect of the state of adoption of Linked Data principles and
technologies in education that is not addressed in this chapter is the way it is
being used. Indeed, we focus here on the data available and the way it is being
modelled, and therefore mostly on the data publication process. The consump-
tion of Web Data for education was actually the main objective of the LinkedUp
project, as illustrated by the LinkedUp Challenge: A series of application devel-
opment competition to encourage the creation of innovative solutions in teaching
and learning through the use of Web Data16
. The result is several dozens of appli-
cations at various stages of maturity. These, as well as other examples, show how
some areas are emerging as the key applications of Linked Data in education,
from the basic management and sharing of data in educational institutions (see
for example [4]) to recommendation (see for example [6]).
On of such area which is generating increasing interest is Learning Analyt-
ics. Learning Analytics is about the processing of data about learners and their
environments for the purpose of understanding and optimising learning (see for
example Ferguson, 2012). A lot of both the research-oriented and the practical
work in this area is dedicated to the methods employed for collecting, analysing,
mining or visualising such data in relation to various levels of models of learning,
from the basic information models used to structure the data, to the cognitive
models that are expected to be reflected in the learners activity patterns found
in the data. It therefore about the way to make sense of raw data in terms of
the learners experience, behaviour and knowledge, and Linked Data could rep-
resent an approach for the collection, integration and dissemination of such data
(see dAquin et al., 2014).
References
1. Alexander, K., Hausenblas, M.: Describing linked datasets - on the design and
usage of void, the vocabulary of interlinked datasets. In: Linked Data on the Web
Workshop (LDOW 09), in Conjunction with 18th International World Wide Web
Conference (WWW 2009) (2009)
2. Brickley, D., Guha, R.V.: RDF vocabulary description language 1.0: RDF schema.
W3C recommendation (2004)
3. Daga, E., d’Aquin, M., Adamou, A., Brown, S.: The open university linked data -
data.open.ac.uk. Semantic Web Journal - Interoperability, Usability, Applicability
(2015)
4. d’Aquin, M.: Putting linked data to use in a large higher-education organisation.
In: Interacting with Linked Data at Extended Semantic Web Conference, ESWC
(2012)
5. d’Aquin, M., Adamou, A., Dietze, S.: Assessing the educational linked data land-
scape. In: ACM Web Science (2013)
14
http://www.lrmi.net.
15
https://www.w3.org/community/opened/.
16
http://linkedup-challenge.org.](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-24-320.jpg)
![Educational Linked Data on the Web -
Exploring and Analysing the Scope
and Coverage
Davide Taibi1()
, Giovanni Fulantelli1
, Stefan Dietze2
,
and Besnik Fetahu2
1
Istituto per le Tecnologie Didattiche, Consiglio Nazionale delle Ricerche,
Palermo, Italy
{davide.taibi,giovanni.fulantelli}@itd.cnr.it
2
L3S Research Center, Hannover, Germany
{dietze,fetahu}@l3s.de
Abstract. Throughout the last few years, the scale and diversity of datasets
published according to Linked Data (LD) principles has increased and also led to
the emergence of a wide range of data of educational relevance. However, suffi-
cient insights into the state, coverage and scope of available educational Linked
Data seem still missing. In this work, we analyse the scope and coverage of
educational linked data on the Web, identifying the most significant resource types
and topics and apparent gaps. As part of our findings, results indicate a prevalent
bias towards data in areas such as the life sciences as well as computing-related
topics. In addition, we investigate the strong correlation of resource types and
topics, where specific types have a tendency to be associated with particular types
of categories, i.e. topics. Given this correlation, we argue that a dataset is best
understood when considering its topics, in the context of its specific resource
types. Based on this finding, we also present a Web data exploration tool, which
builds on these findings and allows users to navigate through educational linked
datasets by considering specific type and topic combinations.
Keywords: Dataset profile Linked data for education Linked data explorer
1 Introduction
The diversity of datasets published according to Linked Data (LD) [5–7] principles has
increased in the last few years and also led to the emergence of a wide range of data of
educational relevance [18]. These include open educational resources metadata, sta-
tistical data about the educational sector, video lecture metadata or university data
about courses, research or experts [2]. Initial efforts to collect and catalogue such
datasets have been made through initiatives such as the LinkedUp Data Catalog1
or
related community initiatives2
.
1
http://data.linkededucation.org/linkedup/catalog/.
2
These include, for instance, http://linkededucation.org, http://linkeduniversities.org or the recently
established W3C Community Group on Open Linked Education (http://www.w3.org/community/
opened/).
© Springer International Publishing Switzerland 2016
D. Mouromtsev and M. d’Aquin (Eds.): Open Data for Education, LNCS 9500, pp. 16–37, 2016.
DOI: 10.1007/978-3-319-30493-9_2](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-26-320.jpg)
![However, the state, coverage and scope of available educational Linked Data have
not been widely investigated. Here, in particular questions about the represented
resource types, such as, resource metadata or information about organisations or
people, and topics are of crucial relevance to shape a better understanding about the
state of educationally relevant Linked Data on the Web. Also identifying a dataset
containing resources related to a specific topic is, at present, a challenging activity.
Moreover, the lack of up-to-date and precise descriptive information has exacerbated
this challenge. The mere keywords-based classification derived from the description of
the dataset owner is not sufficient, and for this reason, it is necessary to find new
methods that exploit the characteristics of the resources within the datasets to provide
useful hints about topics covered by datasets and their subsequent classification.
In this direction, authors in [1, 3] proposed an approach to create structured
metadata to describe a dataset by means of topics, defined as DBpedia categories,
where a weighted graph of topics constitutes a dataset profile. Profiles are created by
means of a processing pipeline3
that combines techniques for datasets resource sam-
pling, topic extraction and topic ranking. Topics have been extracted by using named
entity recognition (NER) techniques, where topics are ranked, respectively weighted,
according to their relevance using graph-based algorithms such as PageRank, K-Step
Markov, and HITS.
The limitations of such an approach are related mainly to the following aspects.
First, the meaning of individual topics assigned to a dataset can be highly dependent on
the type of resources they are attached to. Also, the entire topic profile of a dataset is
hard to interpret if categories from different types are considered at the same time. As
an example of the first issue, the same category (e.g. “Technology”) might be asso-
ciated to resources of very different types such as “video” (e.g. in the Yovisto Datset4
)
or “research institution”(e.g. in the CNR dataset5
). Concerning the second issue, the
single topic profile attached for instance to bibliographic datasets (such as: the LAK
dataset6
or Semantic Web Dog Food7
) - in which people (“authors”), organisations
(“affiliations”) and documents (“papers”) are represented– is characterized by the
diversity of its categories (e.g. DBpedia categories: Scientific_disciplines, Data_man-
agement Information_science but also Universities_by_country, Universities_and_
colleges). Indeed, classification of datasets in the LD Cloud is highly specific to the
resource types one is looking at. While one might be interested in the classification of
“persons” listed in one dataset (for instance, to learn more about the origin countries of
authors in DBLP), another one might be interested in the classification of topics
covered by the documents (for instance disciplines of scientific publications) in the
very same dataset.
3
http://data-observatory.org/lod-profiles/profiling.htm.
4
http://www.yovisto.com/.
5
http://data.cnr.it/.
6
http://lak.linkededucation.org.
7
http://data.semanticweb.org.
Educational Linked Data on the Web - Exploring 17](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-27-320.jpg)
![In this paper, we aim at providing a systematic assessment of educational Linked
Data which consider both, represented topics as well as resource types and their cor-
relations. Questions of interest are:
1. Which types and topics are covered by existing educational Linked Data?
2. What are the central topics covered for particular types (e.g. Open Educational
Resources metadata)?
3. Are certain topics underrepresented for certain types, or vice versa?
The approach we propose overcomes the limitations described above by consid-
ering the topic profiles defined in [3] in the context of the resource types they are
associated with. However, the schemas adopted by the datasets of the LD cloud are
heterogeneous, thus making it difficult to compare the topic profiles across datasets.
While there are many overlapping type definitions representing the same or similar real
world entities, such as “documents”, “people”, “organization”, type-specific profiling
relies on type mappings to improve the comparability and interpretation of types and
consequently, profiles. For this aim the explicit mappings and relations declared within
specific schemas (for instance, foaf:Person being a subclass of foaf:Agent) as well as
across schemas (for instance through owl:equivalentClass or rdfs:subClassOf proper-
ties) are crucial.
While relying on explicit type mappings, we have based our work on a set of
datasets where explicit schema mappings are available from earlier work [2]. This
includes education-related datasets identified by the LinkedUp Catalog8
in combination
with the dataset profiles generated by the Linked Data Observatory9
. While the latter
provides topic profiles for the majority of LOD datasets, the LinkedUp Catalog con-
tains explicit schema mappings which were manually created for the most frequent
types in the LinkedUp Catalog.
The next Section provides a broad overview on the educational Linked Data from a
perspective that highlights the relations with the Open Educational Resource world;
then, we provide a thorough state of the art assessment of the coverage and scope of
educational Linked Data in Sect. 3, which answer aforementioned questions. In
addition, we introduce an interactive explorer of educational Linked Data, in Sect. 4,
which aims at providing a resource type-specific view on categories associated with the
datasets in the LinkedUp Catalog.
2 Resources for Education: Linked Data and OER
The Semantic Web, and specifically the possibility to publish data on the Web and
connect them through links (i.e. the Linked Data model), represents one of the most
significant evolution of the Internet, after the idea of the Web itself.
8
http://data.linkededucation.org/linkedup/catalog/.
9
http://data-observatory.org/lod-profiles.
18 D. Taibi et al.](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-28-320.jpg)
![From an educational point of view, both the human-readable and navigable
structure of the Web pages and the machine-processable datasets of LD have opened up
incredible potentials for the implementation of new and effective pedagogical para-
digms [4].
The hypertextual organization of information and knowledge of the Web has
influenced not only the ICT-based educational projects worldwide, but also the pub-
lication of traditional school textbooks, where anchor-like notes appear throughout a
book for immediate references to other related concepts.
In general, the more evident opportunity of the Web for education is a very basic
one, yet extremely important for education: the possibility to publish information that
everybody can access and use to develop knowledge.
Some years after the birth of the Web, under the pressure of economical, philan-
thropic and pedagogical emergencies, the idea to exploit materials published on the
Web for educational purposes brought to the development of the Open Educational
Resources (OER) movement. Since then, hundreds of OER repositories have populated
the Web with resources designed for education.
From a pedagogical perspective, OER have solved some critics related to the use of
Web pages for education, such as the lack of a pedagogical structure to present
information, or the difficulties in identifying the pedagogical scope of a resource
published on the Web. However, the OER movement does not exclude more general
resources accessible through the Web, provided that they are included into usage
patterns designed according to pedagogical criteria. Furthermore, the spectrum of OER
is really wide, ranging from resources produced by academic or educational committed
institutions to user- or crowd-generated resources. This variety of OER is reflected in
the many definitions of OER that can be found in the literature [13–17].
In spite of their tremendous influence on education, OER have also shown some
limits; amongst the others:
– The lack of a sole standard for OER and repositories, which has fragmented the
offer of OER on the Web [8];
– The complexity in handling direct links between OER and, consequently, in finding
semantically related resources.
– The impossibility to guarantee metadata interoperability, due to the proliferation of
educational metadata schemas [9];
– The impossibility to deal with the vast availability of education-related data on the
web.
The Linked Open Data model offers new solutions for educational resources, partly
solving some of the OER limits, still representing a paradigm that complements the
OER one, and does not substitute it.
Amongst the OER issues that can be solved by the LOD approach:
– LOD are interlinked by definition; consequently, algorithms can automatically
identify semantically related resources; in the case of OER, it was necessary to
develop a semantic layer to describe OERs;
Educational Linked Data on the Web - Exploring 19](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-29-320.jpg)
![– Federated query can be used in order to find resources belonging to distinct datasets;
as far as OER are concerned, this was only possible if OER repositories were
federated, e.g. through the OAI PMH which allows the exposition of metadata
through a common protocol;
– LOD provide the solutions to publish education-related data on the web.
For these reasons, the interest of the educational community in LOD has developed
over the years, even sustained by the growing availability of resources published in the
Linked Data format, which has raised from 12 in 2007 up to 570 in 201410
.
The first applications of Linked Data to education focused on the potentials of LD
to solve interoperability issues in the field of TEL (Technology Enhanced Learning). In
the mEducator project [12], data from a number of open TEL data repositories has been
integrated, exposed and enriched by following LD principles. Afterwards, more and
more attention has been paid to the increasing availability of datasets on the Web, and
particularly to the presence of educational information in the linked data landscape.
The LinkedUp project has explicitly aimed at the educational exploitation of
Linked Open Data, and has distinguished two types of linked datasets: datasets directly
related to educational material and institutions, including information from open
educational repositories and data produced by universities; datasets that can be used in
teaching and learning scenarios, while not being directly published for this purpose.
Therefore, the approach followed by the LinkedUp project enhances the general
principle of the OER movement that not only resources explicitly developed for
educational purposes can be used in educational patterns.
From one hand, this is an essential advantage of open education in general; how-
ever, it amplifies some drawbacks that could hinder the potentials of LOD in education:
– Which datasets and resources can be employed in educational contexts? A similar
challenge has been already addressed in the OER world. However, this task presents
a higher degree of complexity for LOD, since the OER movement focuses on the
development of content on pedagogical principles, while generally there is no
pedagogical theory behind the publication of a dataset, and classifying them
becomes more complicated.
– How datasets (and their resources) should be described in order to facilitate their
search (and pedagogical exploitation)? This issue shows one of the main difference
between OERs and LOD. While a bad-described OER can be easily visited by the
end user in order to check if it is suitable for a specific educational project, a
bad-described dataset can be hardly analysed by the end user, and the risk that the
dataset will be ignored is very high.
For these reasons, specific classification mechanisms as the ones described in this
chapter, which highlight the key elements of a dataset, together with search tools based
on the, are extremely important to fully exploit the potentials of LOD in education.
10
source: http://lod-cloud.net/.
20 D. Taibi et al.](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-30-320.jpg)
![3 Analysing the Coverage of Educational Linked Data
In this section, we present the actual analysis of educational Linked Datasets on the
Web, taking into account both topics as well as resource types.
3.1 Data and Method
Topic annotations are provided in the form of DBpedia categories for the majority of
LD datasets, available from the topic profiles11
dataset, further described in [3]. A topic
profile connects a dataset with the topics extracted from the analysis of resource
samples. Since topics are ranked, a topic profile can be seen as a weighted dataset-topic
graph. As such a, topic profile provides a comprehensive overview of the topic cov-
erage of individual datasets. Analysed across a specific set of datasets - as carried out in
this work - topic profiles provide insights into the coverage of such a set of datasets.
While topic annotations are obtained from analysing resources of a particular type,
the semantics of the topic can best be interpreted when considering the type of the
resource. As an example, if the topic “Biology” is associated to a resource of type foaf:
Document, for instance, a scholarly publication, it indicates that this particular resource
is related to biological aspects. In case the “Biology” topic is associated to a foaf:
Organization resource, it is likely referred to a Biology department of a university. Next
to such differences in interpreting topics, the nature of DBpedia categories also differs
significantly across different types. For instance, while actual document-related types
usually are related to topics which indicate some form of subject or domain (such as
“Biology”), resources which represent some notion of organisation or person usually are
characterised through some broader categorisations, such as “Academic_institutions” or
“People_from_Athens”. These fundamental differences are important to understand the
nature of dataset topic profiles and to motivate our adopted methodology.
Since our work considers the investigation of both, topics and types, we use as
additional data source the LinkedUp Catalog8
. Our research investigates 21 datasets,
which is precisely the set of datasets existing in both collections the LinkedUp Catalog
and the Dataset Topic Profiles, as only for these both topic profile and resource type
mapping annotations were available. The complete list of selected datasets is shown in
Table 1. As explained by Fetahu et al. in [3], topic profiles are generated based on
resource samples, where the applied sampling strategies did take into account factors
such as the population size of respective types leading to different sample sizes across
different datasets. Table 1 indicates both the total amount of data and the characteristics
of the automatically computed sample.
The analysis of the relationships between datasets, topics and resource types -
aimed at providing a response to the research questions posed above - has been
undertaken exploiting network analysis theories and methods. Indeed, the connections
between the three investigated notions can be represented by networks, in which
11
http://data.l3s.de/dataset/linked-dataset-profiles.
Educational Linked Data on the Web - Exploring 21](https://image.slidesharecdn.com/7157-250427090603-6faae24a/85/Open-Data-for-Education-Linked-Shared-and-Reusable-Data-for-Teaching-and-Learning-1st-Edition-Dmitry-Mouromtsev-31-320.jpg)
Ad
More Related Content
Similar to Open Data for Education Linked Shared and Reusable Data for Teaching and Learning 1st Edition Dmitry Mouromtsev (20)
Recently uploaded (20)
Ad
Open Data for Education Linked Shared and Reusable Data for Teaching and Learning 1st Edition Dmitry Mouromtsev
- 1. Open Data for Education Linked Shared and Reusable Data for Teaching and Learning 1st Edition Dmitry Mouromtsev download https://textbookfull.com/product/open-data-for-education-linked- shared-and-reusable-data-for-teaching-and-learning-1st-edition- dmitry-mouromtsev/ Download more ebook from https://textbookfull.com
- 2. We believe these products will be a great fit for you. Click the link to download now, or visit textbookfull.com to discover even more! Development of Linguistic Linked Open Data Resources for Collaborative Data Intensive Research in the Language Sciences 1st Edition Antonio Pareja Lora https://textbookfull.com/product/development-of-linguistic- linked-open-data-resources-for-collaborative-data-intensive- research-in-the-language-sciences-1st-edition-antonio-pareja- lora/ Linked Data for the Perplexed Librarian ALCTS Monograph 1st Edition Scott Carlson https://textbookfull.com/product/linked-data-for-the-perplexed- librarian-alcts-monograph-1st-edition-scott-carlson/ Linked Data Storing Querying and Reasoning 1st Edition Sherif Sakr https://textbookfull.com/product/linked-data-storing-querying- and-reasoning-1st-edition-sherif-sakr/ Teaching physical education for learning Seventh Edition. Edition Rink https://textbookfull.com/product/teaching-physical-education-for- learning-seventh-edition-edition-rink/
- 3. Demystifying big data and machine learning for healthcare 1st Edition Frenzel https://textbookfull.com/product/demystifying-big-data-and- machine-learning-for-healthcare-1st-edition-frenzel/ Azure Data Factory Cookbook - Second Edition Dmitry Foshin https://textbookfull.com/product/azure-data-factory-cookbook- second-edition-dmitry-foshin/ Data Stewardship for Open Science : Implementing FAIR Principles First Edition Mons https://textbookfull.com/product/data-stewardship-for-open- science-implementing-fair-principles-first-edition-mons/ Azure Data Factory Cookbook: Data engineers guide to build and manage ETL and ELT pipelines with data integration , 2nd Edition Dmitry Foshin https://textbookfull.com/product/azure-data-factory-cookbook- data-engineers-guide-to-build-and-manage-etl-and-elt-pipelines- with-data-integration-2nd-edition-dmitry-foshin/ Feature Engineering for Machine Learning and Data Analytics First Edition Dong https://textbookfull.com/product/feature-engineering-for-machine- learning-and-data-analytics-first-edition-dong/
- 4. Dmitry Mouromtsev Mathieu d’Aquin (Eds.) Open Data for Education State-of-the-Art Survey LNCS 9500 123 Linked, Shared, and Reusable Data for Teaching and Learning
- 5. Lecture Notes in Computer Science 9500 Commenced Publication in 1973 Founding and Former Series Editors: Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen Editorial Board David Hutchison Lancaster University, Lancaster, UK Takeo Kanade Carnegie Mellon University, Pittsburgh, PA, USA Josef Kittler University of Surrey, Guildford, UK Jon M. Kleinberg Cornell University, Ithaca, NY, USA Friedemann Mattern ETH Zurich, Zürich, Switzerland John C. Mitchell Stanford University, Stanford, CA, USA Moni Naor Weizmann Institute of Science, Rehovot, Israel C. Pandu Rangan Indian Institute of Technology, Madras, India Bernhard Steffen TU Dortmund University, Dortmund, Germany Demetri Terzopoulos University of California, Los Angeles, CA, USA Doug Tygar University of California, Berkeley, CA, USA Gerhard Weikum Max Planck Institute for Informatics, Saarbrücken, Germany
- 6. More information about this series at http://www.springer.com/series/7409
- 7. Dmitry Mouromtsev • Mathieu d’Aquin (Eds.) Open Data for Education Linked, Shared, and Reusable Data for Teaching and Learning 123
- 8. Editors Dmitry Mouromtsev ITMO University St. Petersburg Russia Mathieu d’Aquin Knowledge Media Institute Milton Keynes UK ISSN 0302-9743 ISSN 1611-3349 (electronic) Lecture Notes in Computer Science ISBN 978-3-319-30492-2 ISBN 978-3-319-30493-9 (eBook) DOI 10.1007/978-3-319-30493-9 Library of Congress Control Number: 2016933113 LNCS Sublibrary: SL3 – Information Systems and Applications, incl. Internet/Web, and HCI © Springer International Publishing Switzerland 2016 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. Printed on acid-free paper This Springer imprint is published by Springer Nature The registered company is Springer International Publishing AG Switzerland
- 9. Preface The amount of open data, including especially linked open data, is constantly increasing in many domains, especially in the public sector. A great number of private and public organizations, institutions, and companies open their data and are interested in efficient solutions for sharing and reuse of published datasets. Obvious benefits come with opening data for end-users, organizations, and developers, by making it easier to find, obtain, and use data independently of their origin, the systems used to produce them, or the applications for which they are intended. This directly connects with the way the areas of learning, teaching, and education are evolving. Indeed, the activity of learning is changing very rapidly, especially through the Web, data, and open tech- nologies. Distance learning is becoming more common, based on openly available educational resources on the Web and the recently appeared massive open online courses (MOOC) both in public higher education institutions and private training centers and organizations. The primary goal of open data in education is therefore to support these changes through new methodologies and technologies that support the sharing and distribution of information about teaching and the subjects of learning. On the practical side, it serves various purposes such as to help teachers to find and create reusable educational materials, to assist students and families in their educational decisions throughout their life, to improve management systems and many others. For this reason the section of educational open data on the Web has expanded with information about courses and educational materials that can be accessed by tools and applications as well as, social and collaborative resources, thus shaping new architectures of open education. The past few years have demonstrated the growing interest in the topic of educational open data and the growth of the community. During five successful editions of the LILE (Linked Learning) workshops, keynotes, paper sessions, and panel discussions have shown the state of the art and progress in practical work with open data in education. A number of initiatives were started including community platforms (such as LinkedUniversities. org), the W3C Open Linked Education Community Group1 , and activities within the Open Knowledge2 and the VIVO platform3 , to name just a few. The goal of this book is therefore to act as a snapshot of current activities, and to share and disseminate the growing collective experience on open and linked data in education. In this volume we bring together research results, studies, and practical endeavors from initiatives spread across several countries around the world. These initiatives are laying the foundations of open and linked data in the education move- ment, and they are leading the way through innovative applications. 1 https://www.w3.org/community/opened/ 2 https://okfn.org/ 3 http://www.vivoweb.org
- 10. The chapters are selected from extended versions of papers presented at an Open Data in Education Seminar4 and the LILE workshops during 2014–20155,6 . They have been chosen to represent the diversity of practices and experiences that exist in the domain, from the researchers, developers, and community leaders who are pioneering the use of open and linked data in education. In the first part of this book, two chapters provide different perspectives on the current state of the use of linked and open data in education, including the use of technology and the topics that are being covered. The second part is to be considered the core of this book as it focuses on the specific, practical applications that are being put in place to exploit open and linked data in education today. In these four chapters, applications are presented ranging from the set- up of open data platforms in educational institutions, to supporting specific learning activities through the use of online, open data. Finally, a key element of the evolving world of open data is to ensure the skills and ability to use such data are there. We therefore focus in the three last chapters of this book on the other side of open and linked data in education: on teaching the technology and practices so they can be widely applied, and on the community of practitioners pushing these practices forward. We assume the readers of this book are reasonably familiar with modern educational technologies and Web standards (including basics of the Semantic Web). The chapters will be of interest, to varying extents, to academic heads and managers; educators, teachers, and tutors, and start-ups in education; library staff; postgraduates; technology researchers and professionals; as well as students and learners who are keen to better understand how the technologies of the Web and linked data can be applied to support progress in learning and education. We acknowledge all the contributors and those who spent time on reviewing chapters and making critical comments and fruitful discussions. First of all we want to thank the members of numerous projects that have supported the development of the works presented in this book, including in particular the LUCERO project, the Lin- kedUp support action, the VIVO project, and some others. We also thank the funders of these projects, as well as our universities and organizations, especially the Open University and ITMO University that provided the environment for such projects to develop. We also want to thank all the members of the various communities dedicated to making open data in education a reality, including the W3C Open and Linked Education community group, the Open Knowledge Open Education Group, LinkedUniversities.org, and LinkedEducation.org. Finally, we thank our families, friends, and colleagues for their support and positive encouragement. January 2016 Dmitry Mouromtsev Mathieu d’Aquin 4 https://linkededucation.wordpress.com/events/open-data-in-education-seminar-st-petersburg/ 5 https://linkededucation.wordpress.com/events/lile2014/ 6 https://lile2015.wordpress.com/ VI Preface
- 11. Contents State of Open and Linked Data for Education On the Use of Linked Open Data in Education: Current and Future Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Mathieu d’Aquin Educational Linked Data on the Web - Exploring and Analysing the Scope and Coverage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Davide Taibi, Giovanni Fulantelli, Stefan Dietze, and Besnik Fetahu Applications of Open and Linked Data in Education ECOLE: An Ontology-Based Open Online Course Platform . . . . . . . . . . . . . 41 Vladimir Vasiliev, Fedor Kozlov, Dmitry Mouromtsev, Sergey Stafeev, and Olga Parkhimovich Use of Semantic Web Technologies in the Architecture of the BBC Education Online Pages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Eleni Mikroyannidi, Dong Liu, and Robert Lee Towards a Linked and Reusable Conceptual Layer Around Higher Education Programs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Fouad Zablith Collaborative Authoring of OpenCourseWare: The Best Practices and Complex Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Darya Tarasowa and Sören Auer Teaching (with) Open and Linked Data Teaching Linked Open Data Using Open Educational Resources. . . . . . . . . . 135 Alexander Mikroyannidis, John Domingue, Maria Maleshkova, Barry Norton, and Elena Simperl On Some Russian Educational Projects in Open Data and Data Journalism. . . 153 Irina Radchenko and Anna Sakoyan The Open Education Working Group: Bringing People, Projects and Data Together. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Marieke Guy Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
- 12. State of Open and Linked Data for Education
- 13. On the Use of Linked Open Data in Education: Current and Future Practices Mathieu d’Aquin(B) Knowledge Media Institute, The Open University, Walton Hall, Milton Keynes, UK [email protected] Abstract. Education has often been a keen adopter of new informa- tion and communication technologies. This is not surprising given that education is all about informing and communicating. Traditionally, edu- cational institutions produce large volumes of data, much of which is publicly available, either because it is useful to communicate (e.g. the course catalogue) or because of external policies (e.g. reports to funding bodies). Considering the distribution and variety of providers (universi- ties, schools, governments), topics (disciplines and types of educational data) and users (students, teachers, parents), education therefore repre- sents a perfect use case for Linked Open Data. In this chapter, we look at the growing practices in using Linked Open Data in education, and how this trend is opening up opportunities for new services and new scenarios. Keywords: Linked data · Semantic web · Education · Learning 1 Why Using Linked Data in Education Traditionally, educational institutions produce large volumes of data, much of which is publicly available, either because it is useful to communicate (e.g., the course catalogue) or because of external policies (e.g., reports to funding bodies). In this context, open data has an important role to play. Implementing open data through Linked Data technologies can be summarized as using the web both as a channel to access data (through URIs supporting the delivery of structured information) and as a platform for the representation and integration of data (through creating a graph of links between these data URIs). Considering the distribution and variety of providers (universities, schools, governments), topics (disciplines and types of educational data) and users (students, teachers, parents), education also represents a perfect use case for Linked Open Data [7]. Indeed, the basic idea of Linked Data [9] is to use the architecture of the Web to share, distribute and interconnect data from various origins into a common, online environment. It is based on the basic principle that raw data objects are identified and accessible, similarly to webpages, through Web addresses (URIs), that deliver the information in a structured, processable and linkable way. c Springer International Publishing Switzerland 2016 D. Mouromtsev and M. d’Aquin (Eds.): Open Data for Education, LNCS 9500, pp. 3–15, 2016. DOI: 10.1007/978-3-319-30493-9 1
- 14. 4 M. d’Aquin This approach has been very successful in the last few years, especially as a base method for the publication of open data on the Web. Linked Data has been adopted by government agencies in several countries (prominently, in the UK and the US) for transparency and public information purposes, by cultural heritage institutions such as libraries and museum to provide more processable and integrated information about their collections (see the Europeana project1 for example, or the British Museum Collection2 ), by companies in publishing (for example at Nature, or Elsevier), broadcasting (for example at the BBC), or retail (for example at BestBuy). As we will see later in this chapter, there is a growing trend in the use of Linked Data specifically for education, with uni- versities in particular making their public information (academic programmes, research outputs, facilities, etc.) available as linked data on the Web (see for example LinkedUniversities.org). 2 Linked Data - In More Details The foundation of the Web is that it is a network of documents connected by hyperlinks. Each document is identified by a Web address, a URI, and might rep- resent a document which content is encoded using a standard, universally read- able format (most commonly HTML). The foundation of Linked Data is that data objects on the Web are identified, similarly to documents, by URIs. The represen- tation of the data – i.e. the information associated with a data object – is then represented by Web links, which can themselves be characterised by URIs. This makes it possible to represent information in such a way that it is materialised as a graph, where nodes are URIs or literal data values (strings, numbers) and the edges are links between them. For example, a university like The Open University3 publishes information about the courses it offers through its website, as well as using linked data [3]. It achieves that through assigning to every course a dedicated URI that acts both as an identifier for the course on the Web, and as a way to address information about this course. For example, http://data.open.ac.uk/course/aa100 is the URI for the course with code AA100, which is an undergraduate (level 1) course in Arts and Humanities, entitled “The arts past and present”. Through the links between this URI and others, information about this course is being represented regarding the topics and description of the course, where it is available, how it is assessed, what course material and open educational resources relate to it, etc. (see Fig. 1). While most of the other data objects it relates to are also identified by URIs within the domain of the Open University, it is important to remark here that it links to other data sources, such as the UK government’s information about The Open University or information provided by the Geonames platform about 1 http://www.europeana.eu/. 2 http://collection.britishmuseum.org/. 3 http://www.open.ac.uk.
- 15. On the Use of Linked Open Data in Education 5 Fig. 1. Extract of the Linked Data (RDF) representation of the course AA100 “The Arts Past and Present” at The Open University (from data.open.ac.uk). the countries in which the course is available. This demonstrates how, from these basic principles, information originating from widely different systems and sources can be seamlessly integrated. Following the base principles described above, the most basic technology employed to implement linked data is a web-enabled, graph-based data rep- resentation language: RDF (Resource Description Framework [10]). RDF is in principle related to XML, but dedicated to the representation of graphs where nodes are URIs or literal values, and edges are links labelled by URIs. It has different syntaxes, including an XML-based one, but also others based on listing the triples [subject, predicate, objects] forming the links in the data. Another important component of the technological stack for linked data is the one of vocabularies. Indeed, it is important that data should be shareable and reusable in a common way across sources and systems. To address that, languages such as RDF-Schema [2] and OWL (the Web Ontology Language [11]) allow one to define the types of objects that can be encountered in the data (the classes, e.g. Course, Person, Country, etc.), as well as the types of relationships that connect these types of objects (the properties, e.g. location, title, employer, author, etc.). Finally, another important element of linked data is the way in which, still relying uniquely on the basic mechanisms of the Web, the data can be consumed. As we already mentioned, URIs on linked data can be requested to obtain RDF (most often in its XML syntax). When more flexibility is required, many of the existing linked data sources offer data endpoint using the standard querying language and protocol for RDF/Linked Data: SPARQL [12]. Briefly, SPARQL is both a query language made explicitly to fit the graph data model of RDF, and a Web protocol dictating the way in which a SPARQL endpoint should be accessed and queried on the Web. For example, the query:
- 16. 6 M. d’Aquin select distinct ?course ?title where { ?course a http://purl.org/vocab/aiiso/schema#Module. ?course http://purl.org/dc/elements/1.1/title ?title. ?course http://data.open.ac.uk/saou/ontology#isAvailableIn http://sws.geonames.org/2328926/. ?course http://purl.org/dc/elements/1.1/subject http://data.open.ac.uk/topic/computing } limit 200 returns the courses (URIs and titles) in computing and IT offered by the Open University and that are available in Nigeria, when executed on the Open Univer- sitys SPARQL endpoint4 . Accessing such a SPARQL endpoint does not require any specific API or library, but is achieved using standard HTTP requests. The query above can therefore also be shared via a standard Web link5 . 3 The Adoption of Linked Data in Education As described above, the elementary principle of the Linked Data of using the Web as a data modelling and access mechanism makes it effective to share and connect information from various sources. This is a property that many institu- tions have already started to exploit, and that is well aligned with the objective of many educational initiatives, especially related to open education: To dissem- inate knowledge resources and enable learning in a connected and global way. In this section, we therefore review the current adoption of these principles and technologies in the area of education, to understand how much this has already happened, and conclude in the next section with views on the next steps and the future of education with open, linked data. We start our analysis with the LinkedUp project6 . Indeed LinkedUp was a European project with the explicit objective to push forward the adoption of Web Data in Education. To support achieving this goal, the project developed a catalogue of education-related Linked Data sources that has grown to several dozen datasets in the last couple of years. Our methodology therefore relies on analysing the content of the LinkedUp Catalogue of Educational Datasets in order to understand the way in which Linked Data has been applied for education already, and what we can expect to happen in the future in this area. 4 http://data.open.ac.uk/query. 5 http://data.open.ac.uk/sparql?query=select%20distinct%20%3Fcourse%20%3Ftitle %20where%20%7B%3Fcourse%20a%20%3Chttps%3A%2F%2Fptop.only.wip.la%3A443%2Fhttp%2Fpurl.org%2Fvocab%2 Faiiso%2Fschema%23Module%3E.%20%3Fcourse%20%3Chttps%3A%2F%2Fptop.only.wip.la%3A443%2Fhttp%2Fpurl. org%2Fdc%2Felements%2F1.1%2Ftitle%3E%20%3Ftitle.%20%3Fcourse%20%3 Chttps%3A%2F%2Fptop.only.wip.la%3A443%2Fhttp%2Fdata.open.ac.uk%2Fsaou%2Fontology%23isAvailableIn%3E%20 %3Chttps%3A%2F%2Fptop.only.wip.la%3A443%2Fhttp%2Fsws.geonames.org%2F2328926%2F%3E.%20%3Fcourse%20% 3Chttps%3A%2F%2Fptop.only.wip.la%3A443%2Fhttp%2Fpurl.org%2Fdc%2Felements%2F1.1%2Fsubject%3E%20%3 Chttps%3A%2F%2Fptop.only.wip.la%3A443%2Fhttp%2Fdata.open.ac.uk%2Ftopic%2Fcomputing%2526it%3E%7D%20 limit%20200. 6 http://linkedup-project.eu.
- 17. On the Use of Linked Open Data in Education 7 3.1 The LinkedUp Project and the LinkedUp Catalogue of Educational Datasets The LinkedUp Project (Linking Web data for education [8]) was an EU FP7 Coordination and Support Action running from November 2012 to November 2014 which looked at issues around open data in education, with the aim of push- ing forward the exploitation of the vast amounts of public, open data available on the Web. The project comprised six pan-European consortium partners led by the L3S Research Center of the Gottfried Wilhelm Leibniz Universitt Han- nover and consisting of the Open University UK, the Open Knowledge Foun- dation, Elsevier, the Open Universiteit Nederland and eXact learning LCMS. The project also had a number of associated partners with an interest in the project including the Commonwealth of Learning, Canada, and the Department of Informatics, PUC-Rio, Brazil. To aid awareness and use of open and linked data in education, the project created and has continuously maintained a catalogue and repository of data relevant and useful to education scenarios. The goal of the LinkedUp Dataset Catalog (or Linked Education Cloud7 ) is to collect and make available, ideally in an easily usable form, all sorts of data sources of relevance to education, pro- viding a shared, evolving resource for the community interested in Web data for education (see Fig. 2). During the project, the technical team has enabled and encouraged content- and data-providers to contribute new material to the LinkedUp Dataset Catalog through a series of hands-on workshops and the pro- motion of community documentation on LinkedUp tools, workflows and lessons learned. Datahub.io is probably the most popular registry of global catalogues of datasets and forms the heart of the Linked Open Data cloud. In the interest of integrating with other ongoing open data effort, rather than developing in isolation, the LinkedUp Data Catalog has been created as part of Datahub.io. It takes the form of a community group in which any dataset can be included, provided that it is relevant, and the datasets in this group are also visible glob- ally on the Datahub.io portal. Every dataset is described with a set of basic metadata and assigned resources. This makes it possible to search for datasets and employ faceted browsing of the results both globally or specifically in the Linked Education Cloud. For example, one could search for the word ‘univer- sity’ in the Linked Education Cloud, and obtain datasets that explicitly mention ‘university’ in their metadata. These results can be further reduced with filters, for example to include only the ones that provide an example resource in the RDF/XML format. One of the key aspects of the design of the LinkedUp catalogue is that it itself creates a Linked Data resource in addition to the use of Datahub.io. Indeed, once datasets have been identified and registered, basic metadata related to each of them, as well as information about their content, are automatically extracted from Datahub.io and from their SPARQL endpoint. This information is then 7 http://data.linkededucation.org/linkedup/catalog/.
- 18. 8 M. d’Aquin Fig. 2. Screenshots of the Web interface to the LinkedUp Catalogue of Datasets for Education for browsing datasets. represented in RDF using the VoID vocabulary8 [1] and made available, in a Linked Data way, through a SPARQL endpoint. It is this SPARQL endpoint that we use and interrogate to analyse the characteristics of existing datasets for education in the next section. 3.2 The State of Linked Data in Education An initial analysis of an earlier version of the catalogue was shown in [5]. It focused on the connection between datasets through their reuse of common vocabulary elements. The core figures from that paper are reproduced in Fig. 3 below, showing the network of datasets and there partitioning through the com- mon reuse of vocabulary elements, and the most commonly used classes/concepts in these datasets, connected by their co-occurrence. The current version of the LinkedUp catalogue is however much bigger: It references 56 different datasets, each with their own SPARQL endpoint. Datasets are obtained from a variety of sources. As can be seen from Fig. 4 however, they essentially originate from either universities publishing their own data, or from repositories of educational or research resources. Government open data also contribute significantly to datasets related to education, with for example statistics about the registration and results of educational institutions. A simple aspect one might want to look at when analysing datasets about education from the LinkedUp Catalogue is the variety of sizes that the datasets represent. Each dataset might in particular be divided into multiple sub-graphs, 8 http://www.w3.org/TR/void/.
- 19. On the Use of Linked Open Data in Education 9 Fig. 3. Dataset (RDF graphs) connected by their reuse of common classes, and common concepts (classes) connected by their co-occurrence (from [5]). Fig. 4. Number of datasets from different areas in the LinkedUp Data Catalogue. which might relate to different topics or originate from different sources. As shown in Fig. 5 (which only includes datasets with more than one sub-graph) the number of graphs included in each dataset can vary enormously (from only 1, to several thousands) depending on the way the dataset has been designed and constructed. For example, the biggest one in number of graphs from Fig. 5 (SEEK-AT-WD) is constituted through crowdsourcing, and assigns a different graph to each contribution. Some universities would include in one graph all the information about all the courses they offer, while others might create a graph for the representation of each course. As can be seen however, besides datasets with very large numbers of graphs, or small datasets focusing on a small number of topics, most datasets are structured using 10 to 100 graphs corresponding to different aspects of the data (e.g. modules, resources, people, facilities, etc.). For information, the chart in Fig. 5 is generated from the results of the fol- lowing query on the SPARQL endpoint of the LinkedUp catalogue9 : 9 http://data.linkededucation.org/linkedup/catalog/sparql/.
- 20. 10 M. d’Aquin Fig. 5. Number of RDF graphs for datasets with more than one graph (log scale). select distinct ?t (count(?sg) as ?n) where { graph http://data.linkededucation.org/linkedup/catalogue/void { ?d a http://rdfs.org/ns/void#Dataset. ?d http://rdfs.org/ns/void#sparqlEndpoint ?x. ?d http://purl.org/dc/terms/title ?t. ?d http://rdfs.org/ns/void#subset ?sg }} group by ?t order by desc(?n) Similarly we can look at the size of dataset through comparing the number of classes and properties they use. To an extent, the number of classes gives an idea of the variety of the dataset, while the number of properties indicates a notion of richness. Figure 6 shows the number of classes and properties of each dataset that refer to at least 1 class in any of their graphs. Once again, it is clear that there is a wide variety across the datasets of the LinkedUp Catalogue. Several datasets cover information about a very small number of classes (sometimes only one), meaning that the focus on a specific and restricted type of data objects (for example educational resources). Amongst these focused datasets, some still use a comparatively large number of properties, indicating that the information available about each data object in those datasets can be expected to be rich. In the other end of the spectrum are datasets with a very large number of classes, which can include dataset representing a thesaurus or classification, where each topic is a class, or others that generate/use very granular classes to represent the different types of objects they represent. To generate the data at the basis of Fig. 6, we used the following SPARQL query to the SPARQL endpoint of the LinkedUp Data Catalogue:
- 21. On the Use of Linked Open Data in Education 11 Fig. 6. Number of classes and number of properties for datasets of the LinkedUp Data Catalogue (log scale). select distinct ?t (count(distinct ?cp) as ?nc) (count(distinct ?pp) as ?np) where { graph http://data.linkededucation.org/linkedup/catalogue/void { ?d a http://rdfs.org/ns/void#Dataset. ?d http://rdfs.org/ns/void#sparqlEndpoint ?x. ?d http://purl.org/dc/terms/title ?t. {{?d http://rdfs.org/ns/void#subset ?sg. ?sg http://rdfs.org/ns/void#classPartition ?cp. ?sg http://rdfs.org/ns/void#propertyPartition ?pp} UNION {?d http://rdfs.org/ns/void#classPartition ?cp. ?d http://rdfs.org/ns/void#propertyPartition ?pp }} }} group by ?t order by desc(?nc) To really understand the way Linked Data is used to represent data for edu- cation, a possible way is to consider an overview of the kind of content they consider. This can especially be done through looking at the types of the data objects that they include, i.e. the classes that they employ to model the data. Looking at Fig. 7, it is interesting to see that, amongst the most popular classes in the datasets, the first is the one used to model people in the FOAF vocabulary10 . 10 http://xmlns.com/foaf/spec/.
- 22. 12 M. d’Aquin Indeed, it appears that many of the educational datasets put a strong emphasis of the way people are involved in education, considering in particular university staff and the way they relate to the educational institutions and organisations they are working with (represented, a bit below in the list, by the Organization and Institution in the FOAF and AIISO11 vocabularies respectively). Unsur- prisingly too, several of the most popular classes relate to the formats in which the data is modeled, including RDF, OWL and DataCube. Again unsurprisingly considering the many datasets originating from repositories (as shown in Fig. 4), most of the remaining classes in Fig. 7 relate to different forms of educational resources or resources that can be used for education, including Document (from FOAF), Article and Book (from the BIBO ontology12 ). Fig. 7. 20 most common classes amongst the datasets in the LinkedUp Data Catalogue (in number of datasets). Similarly to classes, the most common properties used in the datasets is an indication of the focus of the content of datasets included in the LinkedUp catalogue. They however give a slightly different picture, as they do not indicate what kind of objects are represented in the data, but what are the dimensions, attributes or indicators most commonly used to describe them. As can be seen in Fig. 8, besides the properties associated with data formats (RDF, etc.), the majority of the most popular properties relate to the modelling of basic metadata attributes of resources, with the Dublin Core vocabulary13 (for title, description and author for example) as well as to the authors of such resources (for example, the property creator from Dublin Core). Following this, and the conclusion from Fig. 7 that many datasets describe people, we can find amongst the most popular properties also the ones to describe the basic contact information of people, including names, homepages, etc. The query at the basis of Fig. 7 is described below, and can be straightfor- wardly adapted to obtain the data at the basis of Fig. 8. 11 http://vocab.org/aiiso/. 12 http://bibliontology.com/. 13 http://dublincore.org/.
- 23. On the Use of Linked Open Data in Education 13 Fig. 8. 40 most common properties amongst the datasets of the LinkedUp Data Cat- alogue (in number of datasets). select distinct ?c (count(distinct ?d) as ?n) where { graph http://data.linkededucation.org/linkedup/catalogue/void { ?d a http://rdfs.org/ns/void#Dataset {{?d http://rdfs.org/ns/void#subset ?sg. ?sg http://rdfs.org/ns/void#classPartition ?cp. ?cp http://rdfs.org/ns/void#class ?c} UNION {?d http://rdfs.org/ns/void#classPartition ?cp. ?cp http://rdfs.org/ns/void#class ?c} } } } group by ?c order by desc(?n) limit 20 4 Use and Future of Linked Data in Education The analysis described above gives a general overview of the state of Linked Data in education, of the way it is being used and how it can grow further. Indeed, these datasets represent pioneering initiatives that should carry on growing, and through understanding the expending practices of Linked Data in education within these dataset, other can learn from them and have an easier entry point to join the Linked Education CLoud. A key for this to happen however is for these practices to be better shared. Indeed, another element shown in the analysis above is that many of the initiatives at the origin of the considered datasets have been developed in isolation from each other, with different modeling principles, designs and vocabularies being used. While it is the nature of Linked Data (and to an extent of the Web) that it should allow this kind of distribution, some concertation is still required to ensure that the resulting datasets can be used jointly in a way which is sufficiently cohesive (see [5]). Several initiatives have
- 24. 14 M. d’Aquin emerged that are trying to address this, among which LinkedUniversities.org, LRMI14 and the W3C Open Linked Education community group15 . An important aspect of the state of adoption of Linked Data principles and technologies in education that is not addressed in this chapter is the way it is being used. Indeed, we focus here on the data available and the way it is being modelled, and therefore mostly on the data publication process. The consump- tion of Web Data for education was actually the main objective of the LinkedUp project, as illustrated by the LinkedUp Challenge: A series of application devel- opment competition to encourage the creation of innovative solutions in teaching and learning through the use of Web Data16 . The result is several dozens of appli- cations at various stages of maturity. These, as well as other examples, show how some areas are emerging as the key applications of Linked Data in education, from the basic management and sharing of data in educational institutions (see for example [4]) to recommendation (see for example [6]). On of such area which is generating increasing interest is Learning Analyt- ics. Learning Analytics is about the processing of data about learners and their environments for the purpose of understanding and optimising learning (see for example Ferguson, 2012). A lot of both the research-oriented and the practical work in this area is dedicated to the methods employed for collecting, analysing, mining or visualising such data in relation to various levels of models of learning, from the basic information models used to structure the data, to the cognitive models that are expected to be reflected in the learners activity patterns found in the data. It therefore about the way to make sense of raw data in terms of the learners experience, behaviour and knowledge, and Linked Data could rep- resent an approach for the collection, integration and dissemination of such data (see dAquin et al., 2014). References 1. Alexander, K., Hausenblas, M.: Describing linked datasets - on the design and usage of void, the vocabulary of interlinked datasets. In: Linked Data on the Web Workshop (LDOW 09), in Conjunction with 18th International World Wide Web Conference (WWW 2009) (2009) 2. Brickley, D., Guha, R.V.: RDF vocabulary description language 1.0: RDF schema. W3C recommendation (2004) 3. Daga, E., d’Aquin, M., Adamou, A., Brown, S.: The open university linked data - data.open.ac.uk. Semantic Web Journal - Interoperability, Usability, Applicability (2015) 4. d’Aquin, M.: Putting linked data to use in a large higher-education organisation. In: Interacting with Linked Data at Extended Semantic Web Conference, ESWC (2012) 5. d’Aquin, M., Adamou, A., Dietze, S.: Assessing the educational linked data land- scape. In: ACM Web Science (2013) 14 http://www.lrmi.net. 15 https://www.w3.org/community/opened/. 16 http://linkedup-challenge.org.
- 25. On the Use of Linked Open Data in Education 15 6. d’Aquin, M., Allocca, C., Collins, T.: Discou: A flexible discovery engine for open educational resources using semantic indexing and relationship summaries. In: Demo at International Semantic Web Conference, ISWC (2012) 7. d’Aquin, M., Dietze, S.: Open education: A growing, high impact area for linked open data. ERCIM News, (96) (2014) 8. Guy, M., M., d’Aquin, S. Dietze, H. Drachsler, E. Herder, E. Parodi.: Linkedup: Linking open data for education. Ariadne, (72) (2014) 9. Heath, T., Bizer, C.: Linked Data: Evolving the Web Into a Global Data Space. Synthesis Lectures on the Semantic Web: Theory and Technology, 1st edn. Morgan and Claypool, San Francisco (2011) 10. Klyne, G., Carroll, J.J.: Resource description framework (RDF): Concepts and abstract syntax. W3C recommendation (2006) 11. McGuinness, D.L., Van Harmelen, F.: OWL web ontology language overview. W3C recommendation (2004) 12. PrudHommeaux, E., Seaborne, A.: SPARQL query language for RDF. W3C rec- ommendation (2008)
- 26. Educational Linked Data on the Web - Exploring and Analysing the Scope and Coverage Davide Taibi1() , Giovanni Fulantelli1 , Stefan Dietze2 , and Besnik Fetahu2 1 Istituto per le Tecnologie Didattiche, Consiglio Nazionale delle Ricerche, Palermo, Italy {davide.taibi,giovanni.fulantelli}@itd.cnr.it 2 L3S Research Center, Hannover, Germany {dietze,fetahu}@l3s.de Abstract. Throughout the last few years, the scale and diversity of datasets published according to Linked Data (LD) principles has increased and also led to the emergence of a wide range of data of educational relevance. However, suffi- cient insights into the state, coverage and scope of available educational Linked Data seem still missing. In this work, we analyse the scope and coverage of educational linked data on the Web, identifying the most significant resource types and topics and apparent gaps. As part of our findings, results indicate a prevalent bias towards data in areas such as the life sciences as well as computing-related topics. In addition, we investigate the strong correlation of resource types and topics, where specific types have a tendency to be associated with particular types of categories, i.e. topics. Given this correlation, we argue that a dataset is best understood when considering its topics, in the context of its specific resource types. Based on this finding, we also present a Web data exploration tool, which builds on these findings and allows users to navigate through educational linked datasets by considering specific type and topic combinations. Keywords: Dataset profile Linked data for education Linked data explorer 1 Introduction The diversity of datasets published according to Linked Data (LD) [5–7] principles has increased in the last few years and also led to the emergence of a wide range of data of educational relevance [18]. These include open educational resources metadata, sta- tistical data about the educational sector, video lecture metadata or university data about courses, research or experts [2]. Initial efforts to collect and catalogue such datasets have been made through initiatives such as the LinkedUp Data Catalog1 or related community initiatives2 . 1 http://data.linkededucation.org/linkedup/catalog/. 2 These include, for instance, http://linkededucation.org, http://linkeduniversities.org or the recently established W3C Community Group on Open Linked Education (http://www.w3.org/community/ opened/). © Springer International Publishing Switzerland 2016 D. Mouromtsev and M. d’Aquin (Eds.): Open Data for Education, LNCS 9500, pp. 16–37, 2016. DOI: 10.1007/978-3-319-30493-9_2
- 27. However, the state, coverage and scope of available educational Linked Data have not been widely investigated. Here, in particular questions about the represented resource types, such as, resource metadata or information about organisations or people, and topics are of crucial relevance to shape a better understanding about the state of educationally relevant Linked Data on the Web. Also identifying a dataset containing resources related to a specific topic is, at present, a challenging activity. Moreover, the lack of up-to-date and precise descriptive information has exacerbated this challenge. The mere keywords-based classification derived from the description of the dataset owner is not sufficient, and for this reason, it is necessary to find new methods that exploit the characteristics of the resources within the datasets to provide useful hints about topics covered by datasets and their subsequent classification. In this direction, authors in [1, 3] proposed an approach to create structured metadata to describe a dataset by means of topics, defined as DBpedia categories, where a weighted graph of topics constitutes a dataset profile. Profiles are created by means of a processing pipeline3 that combines techniques for datasets resource sam- pling, topic extraction and topic ranking. Topics have been extracted by using named entity recognition (NER) techniques, where topics are ranked, respectively weighted, according to their relevance using graph-based algorithms such as PageRank, K-Step Markov, and HITS. The limitations of such an approach are related mainly to the following aspects. First, the meaning of individual topics assigned to a dataset can be highly dependent on the type of resources they are attached to. Also, the entire topic profile of a dataset is hard to interpret if categories from different types are considered at the same time. As an example of the first issue, the same category (e.g. “Technology”) might be asso- ciated to resources of very different types such as “video” (e.g. in the Yovisto Datset4 ) or “research institution”(e.g. in the CNR dataset5 ). Concerning the second issue, the single topic profile attached for instance to bibliographic datasets (such as: the LAK dataset6 or Semantic Web Dog Food7 ) - in which people (“authors”), organisations (“affiliations”) and documents (“papers”) are represented– is characterized by the diversity of its categories (e.g. DBpedia categories: Scientific_disciplines, Data_man- agement Information_science but also Universities_by_country, Universities_and_ colleges). Indeed, classification of datasets in the LD Cloud is highly specific to the resource types one is looking at. While one might be interested in the classification of “persons” listed in one dataset (for instance, to learn more about the origin countries of authors in DBLP), another one might be interested in the classification of topics covered by the documents (for instance disciplines of scientific publications) in the very same dataset. 3 http://data-observatory.org/lod-profiles/profiling.htm. 4 http://www.yovisto.com/. 5 http://data.cnr.it/. 6 http://lak.linkededucation.org. 7 http://data.semanticweb.org. Educational Linked Data on the Web - Exploring 17
- 28. In this paper, we aim at providing a systematic assessment of educational Linked Data which consider both, represented topics as well as resource types and their cor- relations. Questions of interest are: 1. Which types and topics are covered by existing educational Linked Data? 2. What are the central topics covered for particular types (e.g. Open Educational Resources metadata)? 3. Are certain topics underrepresented for certain types, or vice versa? The approach we propose overcomes the limitations described above by consid- ering the topic profiles defined in [3] in the context of the resource types they are associated with. However, the schemas adopted by the datasets of the LD cloud are heterogeneous, thus making it difficult to compare the topic profiles across datasets. While there are many overlapping type definitions representing the same or similar real world entities, such as “documents”, “people”, “organization”, type-specific profiling relies on type mappings to improve the comparability and interpretation of types and consequently, profiles. For this aim the explicit mappings and relations declared within specific schemas (for instance, foaf:Person being a subclass of foaf:Agent) as well as across schemas (for instance through owl:equivalentClass or rdfs:subClassOf proper- ties) are crucial. While relying on explicit type mappings, we have based our work on a set of datasets where explicit schema mappings are available from earlier work [2]. This includes education-related datasets identified by the LinkedUp Catalog8 in combination with the dataset profiles generated by the Linked Data Observatory9 . While the latter provides topic profiles for the majority of LOD datasets, the LinkedUp Catalog con- tains explicit schema mappings which were manually created for the most frequent types in the LinkedUp Catalog. The next Section provides a broad overview on the educational Linked Data from a perspective that highlights the relations with the Open Educational Resource world; then, we provide a thorough state of the art assessment of the coverage and scope of educational Linked Data in Sect. 3, which answer aforementioned questions. In addition, we introduce an interactive explorer of educational Linked Data, in Sect. 4, which aims at providing a resource type-specific view on categories associated with the datasets in the LinkedUp Catalog. 2 Resources for Education: Linked Data and OER The Semantic Web, and specifically the possibility to publish data on the Web and connect them through links (i.e. the Linked Data model), represents one of the most significant evolution of the Internet, after the idea of the Web itself. 8 http://data.linkededucation.org/linkedup/catalog/. 9 http://data-observatory.org/lod-profiles. 18 D. Taibi et al.
- 29. From an educational point of view, both the human-readable and navigable structure of the Web pages and the machine-processable datasets of LD have opened up incredible potentials for the implementation of new and effective pedagogical para- digms [4]. The hypertextual organization of information and knowledge of the Web has influenced not only the ICT-based educational projects worldwide, but also the pub- lication of traditional school textbooks, where anchor-like notes appear throughout a book for immediate references to other related concepts. In general, the more evident opportunity of the Web for education is a very basic one, yet extremely important for education: the possibility to publish information that everybody can access and use to develop knowledge. Some years after the birth of the Web, under the pressure of economical, philan- thropic and pedagogical emergencies, the idea to exploit materials published on the Web for educational purposes brought to the development of the Open Educational Resources (OER) movement. Since then, hundreds of OER repositories have populated the Web with resources designed for education. From a pedagogical perspective, OER have solved some critics related to the use of Web pages for education, such as the lack of a pedagogical structure to present information, or the difficulties in identifying the pedagogical scope of a resource published on the Web. However, the OER movement does not exclude more general resources accessible through the Web, provided that they are included into usage patterns designed according to pedagogical criteria. Furthermore, the spectrum of OER is really wide, ranging from resources produced by academic or educational committed institutions to user- or crowd-generated resources. This variety of OER is reflected in the many definitions of OER that can be found in the literature [13–17]. In spite of their tremendous influence on education, OER have also shown some limits; amongst the others: – The lack of a sole standard for OER and repositories, which has fragmented the offer of OER on the Web [8]; – The complexity in handling direct links between OER and, consequently, in finding semantically related resources. – The impossibility to guarantee metadata interoperability, due to the proliferation of educational metadata schemas [9]; – The impossibility to deal with the vast availability of education-related data on the web. The Linked Open Data model offers new solutions for educational resources, partly solving some of the OER limits, still representing a paradigm that complements the OER one, and does not substitute it. Amongst the OER issues that can be solved by the LOD approach: – LOD are interlinked by definition; consequently, algorithms can automatically identify semantically related resources; in the case of OER, it was necessary to develop a semantic layer to describe OERs; Educational Linked Data on the Web - Exploring 19
- 30. – Federated query can be used in order to find resources belonging to distinct datasets; as far as OER are concerned, this was only possible if OER repositories were federated, e.g. through the OAI PMH which allows the exposition of metadata through a common protocol; – LOD provide the solutions to publish education-related data on the web. For these reasons, the interest of the educational community in LOD has developed over the years, even sustained by the growing availability of resources published in the Linked Data format, which has raised from 12 in 2007 up to 570 in 201410 . The first applications of Linked Data to education focused on the potentials of LD to solve interoperability issues in the field of TEL (Technology Enhanced Learning). In the mEducator project [12], data from a number of open TEL data repositories has been integrated, exposed and enriched by following LD principles. Afterwards, more and more attention has been paid to the increasing availability of datasets on the Web, and particularly to the presence of educational information in the linked data landscape. The LinkedUp project has explicitly aimed at the educational exploitation of Linked Open Data, and has distinguished two types of linked datasets: datasets directly related to educational material and institutions, including information from open educational repositories and data produced by universities; datasets that can be used in teaching and learning scenarios, while not being directly published for this purpose. Therefore, the approach followed by the LinkedUp project enhances the general principle of the OER movement that not only resources explicitly developed for educational purposes can be used in educational patterns. From one hand, this is an essential advantage of open education in general; how- ever, it amplifies some drawbacks that could hinder the potentials of LOD in education: – Which datasets and resources can be employed in educational contexts? A similar challenge has been already addressed in the OER world. However, this task presents a higher degree of complexity for LOD, since the OER movement focuses on the development of content on pedagogical principles, while generally there is no pedagogical theory behind the publication of a dataset, and classifying them becomes more complicated. – How datasets (and their resources) should be described in order to facilitate their search (and pedagogical exploitation)? This issue shows one of the main difference between OERs and LOD. While a bad-described OER can be easily visited by the end user in order to check if it is suitable for a specific educational project, a bad-described dataset can be hardly analysed by the end user, and the risk that the dataset will be ignored is very high. For these reasons, specific classification mechanisms as the ones described in this chapter, which highlight the key elements of a dataset, together with search tools based on the, are extremely important to fully exploit the potentials of LOD in education. 10 source: http://lod-cloud.net/. 20 D. Taibi et al.
- 31. 3 Analysing the Coverage of Educational Linked Data In this section, we present the actual analysis of educational Linked Datasets on the Web, taking into account both topics as well as resource types. 3.1 Data and Method Topic annotations are provided in the form of DBpedia categories for the majority of LD datasets, available from the topic profiles11 dataset, further described in [3]. A topic profile connects a dataset with the topics extracted from the analysis of resource samples. Since topics are ranked, a topic profile can be seen as a weighted dataset-topic graph. As such a, topic profile provides a comprehensive overview of the topic cov- erage of individual datasets. Analysed across a specific set of datasets - as carried out in this work - topic profiles provide insights into the coverage of such a set of datasets. While topic annotations are obtained from analysing resources of a particular type, the semantics of the topic can best be interpreted when considering the type of the resource. As an example, if the topic “Biology” is associated to a resource of type foaf: Document, for instance, a scholarly publication, it indicates that this particular resource is related to biological aspects. In case the “Biology” topic is associated to a foaf: Organization resource, it is likely referred to a Biology department of a university. Next to such differences in interpreting topics, the nature of DBpedia categories also differs significantly across different types. For instance, while actual document-related types usually are related to topics which indicate some form of subject or domain (such as “Biology”), resources which represent some notion of organisation or person usually are characterised through some broader categorisations, such as “Academic_institutions” or “People_from_Athens”. These fundamental differences are important to understand the nature of dataset topic profiles and to motivate our adopted methodology. Since our work considers the investigation of both, topics and types, we use as additional data source the LinkedUp Catalog8 . Our research investigates 21 datasets, which is precisely the set of datasets existing in both collections the LinkedUp Catalog and the Dataset Topic Profiles, as only for these both topic profile and resource type mapping annotations were available. The complete list of selected datasets is shown in Table 1. As explained by Fetahu et al. in [3], topic profiles are generated based on resource samples, where the applied sampling strategies did take into account factors such as the population size of respective types leading to different sample sizes across different datasets. Table 1 indicates both the total amount of data and the characteristics of the automatically computed sample. The analysis of the relationships between datasets, topics and resource types - aimed at providing a response to the research questions posed above - has been undertaken exploiting network analysis theories and methods. Indeed, the connections between the three investigated notions can be represented by networks, in which 11 http://data.l3s.de/dataset/linked-dataset-profiles. Educational Linked Data on the Web - Exploring 21
- 32. the elements are nodes and their relationship are edges. Specifically the analysis of the relationships has been conducted by considering: – the network representing the relationships between datasets mediated by categories/topics – the network representing the relationships between datasets mediated by resource types – the network representing the relationships between resource types mediated by categories/topics These networks have been represented by using the Open Source software Gephi12 . Due to the high number of categories connected to certain datasets (as shown in Table 1), dataset profiles have been filtered by selecting for each dataset the top 100 categories with the highest relevance score. Exploiting the insights gained from such networks, we can identify the particular type/topic coverage of educational LD datasets, corresponding gaps, and the correlation of educational resource types and topics. 3.2 Analysing Topic Coverage - the Dataset-Category-Graph Representing datasets and categories, i.e. topics, as a weighted graph allows us to analyse the topic coverage of assessed datasets and their proximity topic-wise. In particular, Table 1. Datasets, resources and resource types Dataset Total data Sampled data #Types #resource # Types #resource # Categ. asn-us 29 7494200 3 10000 2128 Colinda 21 17006 9 1985 479 data-cnr-it 120 485977 7 29768 2702 data-open-ac-uk 134 386291 7 36668 1979 education-data-gov-uk 99 315632 42 18712 2510 educationalprograms_sisvu 27 104238 22 12627 2144 gesis-thesoz 9 48532 4 1176 487 hud-library-usagedata 6 904747 1 10000 2300 l3s-dblp 6 15514 3 9368 943 lak-dataset 14 13688 3 10000 1496 linked-open-aalto-data-service 22 373553 12 17598 1543 Morelab 13 244 9 890 206 open-courseware-consortium-metadata-in-rdf 4 22850 4 29369 2723 organic-edunet 1 11093 1 847 559 Oxpoints 142 73655 30 8649 1529 publications-of-charles-university-in-prague 258 14324 15 658 197 seek-at-wd-ict-tools-for-education-web-share 556 13502 37 9938 1624 unistat-kis-in-rdf-key-information-set-uk-universities 35 371737 9 39684 556 universitat-pompeu-fabra-linked-data 39 5778 13 1617 312 university-of-bristol 15 240179 12 22572 2450 Yovisto 8 549986 8 5605 2122 12 http://gephi.github.io/. 22 D. Taibi et al.
- 33. Another Random Scribd Document with Unrelated Content
- 34. Vagy: Isten vesztene el! Vagy: ördög szaggatna el! Ne mondjad nékiek. De minden te dolgodban Légy szelid jó kedvű; Szódban, gondolatodban Igaz Istenfélő. Igy lészesz az Istennél Mennyben ő Fölségénél, Míg itt lészesz kedves; Halálod óráján is – Sőt halálod után is Mindörökkén fénlő. Ez legyen mastan elég Első tanácsomban; Ez nékem szintén elég Az új házasságban: Ha ezeket cselekszed – Szeretetemet veszed Mig élsz ez világban. Ha penig általhágod: Igy ez jutalmát várod Még idő jártában. – Igy tanítá mátkáját Egy korban egy ifjú Újonnand hozott társát, De vége lőn csak bú; Mert az hitet megszegé És csak tréfának vélé. Az szél már rosszul fú! Házasságot megunta, Sok izben megsiratta – Még most is szomorú.
- 35. * * * S z e n c s e y G y ö r g y dalkönyvéből. JAJ NEKEM SZEGÉNYNEK… – A XVI-ik vagy XVII-ik századból. – Jaj nekem szegénynek Idegen legénynek Ki mast utra indulok; Szokatlan helyekre, Hegyekre völgyekre Immár ma s holnap jutok, Az én asszonkámnak Szép palotájának Füstire csak vigyázok. Szüzek szép virágok Én vagyok példátok – Istennek szolgáljatok, Hogy itt ez világon Az szerelem miatt Tőrben ne akadjatok, Bánat keserüség És siralom miatt Ti el ne hervadjatok. Nohát édes lölköm Kihez hajtsam fejem? Mert nincs nékem szerelmem, – Sok irigyek miatt, Gonosz nyelvek miatt
- 36. Elhagyattaték tűlem; Kegyetlen vadakkal Avvagy farkasokkal Együtt hagyott már engem. Siralmas életem, Szomorú én lölköm, Nincs semmi keservesem: Mert még az vadak is Az oroszlánok is Még ők is szánnak engem. Szép szavú rigócskák Csácsogó szajkócskák Együtt siratnak engem. Adta volna Isten Az te szépségedet Hogy ne láthattam volna, Vagy anyád méhében Soha ez világra Ne születtettél volna. Adta volna Isten: Az te szépségedben Ne részesültem volna. Lölköm szép asszonyom Mondd meg szolgálatom: Hogy ha köllök-é, vagy nem? Mert az hajnalcsillag Mikoron el-föl jün: Fülemüle akkor szól, Az ő szép szavával És szép énekével Hasogatja ő szivét. * * *
- 37. E gyönyörű – bár ugy látszik csonka végü – dalt S z e n c s e y György daloskönyve tartotta fenn. Egész modora, versmértéke, a „csácsogó“, „el-föl-jün“ stb. régiesb kifejezések a Balassa idejébe: a XVI-ik század második felébe, vagy tám legkésőbben a XVII-ik első éveibe helyezik keletét. T. K. BÁTOR BÁR UGY LÉGYEN… – A XVI-ik vagy XVII-ik századból. – Bátor bár ugy légyen mint hozza szerencse: Megnyugodtam rajta – teljék kedve benne! Engedek mindenben csak legyen érdemem, Holtig rabod lészek, mert vettél kedvedben. Óh melly nehéz dolog sokat várakozni – Várakodásának jutalmát nem venni! De csoda az idő: attól is kell várni: Ha magának ember nem akar ártani. Az szerencse ollyan mint az forgó kerek – De vakot ne vessen, mert az koczka megdül. Mert mind éjjel nappal szivem téged óhajt: Kedvedben ez világ most néked nem adhat! Tartson meg az Isten tégedet sokáig, Hogy vígasztalj engem mindenkoron szívem!
- 38. Ajánlom magamat holtiglan hűséggel: Mert feljegyzettelek tégedet szivemben. Nálad minden kedvem vagyon elrejtetve – Kérlek jóakaróm, ne legyen elvesztve! Adja meg az Isten rövidnap azt érnem: Hogy teveled legyen nékem minden kedvem. Éltessen az Isten téged szép Katuskám Kláris ajakiddal zöngő szép madarkám. Se téged, se engem búra ne taszítson – Mint rózsát harmattal szépen megújítson. Irám ez verseket szeretőm kedvéért Az én hozzám való kedves hűségéért. * * * M á t r a y - c o d e x . Ez ének rythmusa szabatos, de rímelése oly kezdetleges, a minőt a XVI. századon innen már nem igen találunk. Az első sorban előjövő „Bátor bár“ sajátságos együtthasználása két egy jelentésü szónak. T. K. SÖRKENJ FÖL ÉN LÖLKÖM… – A XVI-ik vagy XVII-ik századból. – Sörkenj föl én lölköm, Kiálts Istenedhez
- 39. Illyen nagy szükségedben; Minden kétség nélkül Bizzál csak ő benne – Meghallgat kérésedben. Imé Uram Isten Mely nagy sokan vannak. Kik ellenem támadnak; Nincsen nyugodalmam, Minden felől látom Én reám támadtanak. Gyakorta búsulok Ezt mondván óhajtok: Hogy az Isten engemet Szinte elfelejtett Előle elvetett – Hová hajtsam fejemet? Siralmas lefektem, Siralmas felköltöm – Megyek immár Istenem! Vagy ülök vagy állok: Búbánatban vagyok… Jaj már az én életem! Régi ismerőim Sok jó akaróim Engemet megvetettek; Már szegény fejemre Mintegy ellenségre Szintén ugy támadtanak. Ne hagyj én Istenem! Hallgasd meg kérésem, Tekénts reám árvádra;
- 40. Nyisd meg az egeket, Hajts(d) le füleidet Az én imádságomra! Mastani igyemben Keserüségemben Kérlek, vigasztalj engem! Mert tebenned vagyon Én édes Istenem, Hitem és reménségem! Vedd el bánatimat Fordíts(d) siralmimat Kérlek, boldog örömben; Vigságos napokon Engedd hogy láthasson Ez földön életemben! Hallgasd meg kérésem Én édes Istenem Az te áldott Fiadért! Az egy idvőzitő Szép Jézus Christusért, Mi áldott egy Urunkért. Ámen. * * * Szencsey dalkönyvéből. VAGYON-É SZIVEM SZÁNDÉKODBAN…
- 41. – Igen régi. – Vagyon-é szivem szándékodban Hogy béfogadsz engem? Igen is vagyon szivem, lelkem – Csak legyen jó kedved! Add kezemben jobb kezedet – Okossan forgassad én édes violám! Lám megholdult, lám megholdult Én árva fejem néked! De sőt inkább ápolgatnám Ragyogó villogós két szép szemed! * * * E dalocskán, melyet a XVII-ik századbeli M á t r a y - c o d e x b e n találtam, oly ódon szinezet ömlik el, a rimelésnek úgy szólva még csak sejtelme mutatkozik, s a rythmus alakja oly ószerü: miszerint bizton merném állítani, hogy a XVI-ik, vagy talán még a XV-ik századból való, s egykorú lehet a maig is élő gyermekdallal: „ L e n g y e l L á s z l ó j ó k i r á l y u n k …“, a melynek rythmusához a jelen dalocska rythmusa helylyel-közzel sokat hasonlít. T. K.
- 42. Lábjegyzetek. 1) Még ma, Biró Márton veszprémi püspök ideje után is túlnyomólag protestáns helység a Bakonyban, Veszprémhez nem messze. T. K. 2) Azaz: engedelmet adó, engedelemmel teljes. 3) I. János és I. Ferdinánd között. 4) Ugy látszik a töröknek a Zápolya-párt által segitségül hivására czéloz. T. K. 5) Korántsem, távulról sem. 6) V é g h á z , v é g v á r : a törökök ellen fennállott határvár; innét v é g b e l i e k : azon vitézek, kik e várak őrségét képezvén, folytonos csatározásban voltak, s a legjobb hősökké fejlődének. A vég-szó akkortájban az illető várak neve elé iratott, igy találjuk például gyakorta: Vég-Szendrő, Vég-Ónad, Vég-Veszprém, Vég- Ujvár, Vég-Simontornya stb. vára. – T. K. 7) Az az nem veszed Krisztus testének jegyeit: a kenyeret és bort az úrvacsorában. T. K. 8) Semmivé. T. K. 9) Ébredjetek helyett, ódon. T. K. 10) A miket helyett, ószerü.
- 43. 11) Ezen, és a következő több versszak későbbi kéz által kegyeletlenül kitörültetett, úgy hogy csak nagy ügygyel bajjal, nagyító üveggel lehet elolvasni. 12) Breviarium. T. K. 13) Értsd: a pápa; bétakará: betakarítá. 14) Betlehemben, összevonva. 15) Kimutatta magát, kijelentkezett. T. K. 16) Jellemző az akkori protestans magyarok buzgóságára nézve azon hiedelem, mely a hegedős ime verszakából látszik, hogy ők t. i. a törökök bejövetelét, pusztitásait a hitjavitás előtti sötét középkor túlcsapongásai, bűnei miatt való isten-ostorának, büntetésnek tartották. Igy egyesíté a magyar, hazája ügyét még vallásával is mindenha. 17) A bibliában. T. K. 18) Czélzás Szent-László nagy érczlovagszobrára, mely Nagy- Váradon – e püspöki székhelyen – állott, mignem a törökök a várat 1660-ban bevevén, ágyukat öntének belőle. 19) Azaz: kezére ne kerítse Váradot; annálinkább félhettek ettől hegedősünk korában, mivel F r á t e r G y ö r g y váradi püspök, I z a b e l l a királyné mindenható kincstárnoka – mint tudva van, – 1549-ben még a törökkel tartott. 20) Szent-László diszes márvány-koporsóban Váradon feküdt. T. K. 21) Erre vonatkozólag mondja a szerző a kettővel föntebb álló versszakban: „Hiszem az Istent állítod vaknak.“ értvén, hogy Isten helyett Szent-László fejét imádják, melynek immár szemei nincsenek. T. K.
- 44. 22) M i v e l t e k . Mennyivel inkább ráillenék e feddés pénz-kapzsi korunkra! T. K. 23) Pokol módra, gonoszúl. T. K. 24) Korántsem. T. K. 25) Húnn-őseink telepedését érti. 26) Szükség. 27) Az az: keresztyénségben. T. K. 28) Tőlük, általuk. T. K. 29) Ugyanaz. T. K. 30) Mit. 31) Tehát, azért. T. K. 32) Főispánságban. T. K. 33) M e g k é m l e n é k helyett régies, mint föntebb a „mihelyen“ m i h e l y t helyett. T. K. 34) A k i n c s e s epithetonnal hajdanában rendesen Buda, Erdély és Kolozsvár szoktak illettetni. T. K. 35) Azaz: menten, azonnal. T. K.
- 45. 36) N é p b ő l helyett, régies; majd mindíg igy használtatik. 37) N y a r g a l á n a k helyett, régies; néhol „jargalának“ is fordul elé. T. K. 38) Elnézék. 39) Mind-en, azaz valamennyien; túl a Dunán ma is igy használtatik. T. K. 40) Az az megenyhitvén, megvigasztalván. 41) Eleve, eleibe. T. K. 42) K ü l s ő n é p e k : Erdélyen kivüliek, az az magyarországiak. T. K. 43) Az az: a vajdának (urnak) jelenté minden vitéz a maga hű voltát. T. K. 44) Mert hogy helyett, régies. 45) „Az önnen seregét“: azaz a magyar hadat; „közübben“, k ö z ü k b e n vagy k ö z é p b e n helyett. 46) Tartalék. (Reserv corps.) T. K. 47) Előbbre. 48) Mily dolog t ö r t é n t e g y s z e r r e . 49) Egyszerre, együtt; ugyan igy értsd e stropha végsorában is. T. K. 50) Egyetemben megsebesíttetett. T. K.
- 46. 51) Az vagy = avvagy; egyébiránt e sor ugy látszik toldva is van, tán a másoló tévedéséből. T. K. 52) Szarczolása, sarcza, váltságdíja. T. K. 53) G a z d a g helyett, régies. T. K. 54) Lakomájokban. 55) Aztán vagy azonnal helyett, régies. T. K. 56) „Azonkivül“ értelemben. T. K. 57) B e r e n d e z t e helyett, régenten mindíg igy használtatik; például „rendelt seregekkel“ az az: rendezett seregekkel. T. K. 58) Udvarmesterének. T. K. 59) Iskolára. T. K. 60) Czélzás II-ik János királyra. 61) Báthori Zsigmond. 62) Irigyek, kegyetlenek. T. K. 63) Jó hirök; jót beszéljenek róluk. T. K. 64) Értsd végházainknak, azaz végvárainknak. T. K.
- 47. 65) Döltsd = döntsd értelemben. T. K. 66) E b i z n á helyett régies. T. K. 67) Dampierre. 68) Bucquoi. T. K. 69) K ö z v o x - ú l = közszavazattal, általános szavazattal. T. K. 70) Értsd: A mig belőled csak egy ember is létezik. T. K. 71) A p r ó b a régi értelemben = harcz; még a II. Rákóczi Ferencz korabeli vezérek levelezéseiben is majdnem mindíg ez értelemben fordul elő. T. K. 72) Tán: o p i u m ? T. K. 73) Mehmet. T. K. 74) Az az: d e r é k , j e l e s ; mint ma is: szép dolog = jeles dolog. T. K. 75) Szakoly, Szabolcsban. T. K. 76) Azaz: odahagyta izmossága, ereje. T. K. 77) Szép = derék, jeles; mint már elébb eléfordult vala. T. K. 78) Azaz: oly gyenge mint a hárshéj.
- 48. 79) A török czímer félholdja. 80) Rendezett, mai szólam szerént. 81) Muzulmán. T. K. 82) Gr. Z r í n y i Miklóst a hőst és költőt érti. T. K. 83) Az az fölemelé t. i. a dárdát. T. K. 84) Az az: Zrínyi összemarczangoltatását. 85) E várakat akkor török birta. T. K. 86) Az az: valaha, egykor. T. K. 87) II. Rákóczi György erdélyi fejedelemé, ki Gyalunál esett el a török elleni harczban. T. K. 88) Az az: örömlövésekkel megünnepeltetik. T. K. 89) Kisfaludy László e török fogságából később kiszabadult, mert 1704-ben már II. Rákóczi Ferencz hadseregében látjuk őt küzdeni mint ezeres kapitányt. – Septembernek (Kisasszony hava) csak 30 napja van, igy a versszerző, vagy e napon, vagy oct. 1-jén irá költeményét, a dátumban tévedvén. T. K. 90) A Gergely név diminutivuma; Erdélyben ma is széltére használtatik. T. K. 91) Keresett, szerzett. T. K.
- 49. 92) Egyébiránt hogy még az ezredekben rendesen szolgáló katonákat is magok a vezérek szegénylegényeknek nevezték: ezt történelmi kutatásaim folytán számtalan példával bebizonyithatom akár Rákóczi és Bercsényi, akár Bezerédi Imre és Béri Balogh Ádám – e két hires kurucz brigadéros – eredeti leveleiből, ugy egykorú kurucz dalokból is. T. K. 93) Elmulasztottál értelemben, régies. T. K. 94) Az az: szánjad. T. K. 95) Gyászban. T. K. 96) N a g y - u r a m ! Dunántúl maig is divatos megszólítás a kisebbrendü nemeseket illetőleg. T. K. 97) Fönntartott, fölnevelt. T. K. 98) Kezök helyett, régen mindíg igy használák. T. K. 99) Nemzetségem, családom. T. K. 100) Jeles, derék. T. K. 101) Jeles, derék. T. K. 102) Az eredetiben is e magyarázat van hozzá téve: „ P r ó f é t a , v a g y n é z ő .“ Ma: látnok. 103) F e l v o n v a értelemben.
- 50. T. K. 104) Többi. T. K. 105) Rimánkodnak. T. K. 106) Újítsd: az az v i d á m í t s d meg. T. K. 107) Mária-Terézia. T. K. 108) Katonát: azaz lovast; hajdut: azaz gyalogot; a „paraszti renden lévők“ ugynevezett „portális hajdu“-kat – kapuszám szerint kivetett gyalogokat – tartozának kiállítani. T. K. 109) Köpönyeg a b a - p o s z t ó b ó l , mely alatt őseink durva szövetü posztót értettek; a magyar katonaságnak – legalább a kuruczoknak, mint egykorú számadásokból látom – nadrágjaik és köpenyeik ily kelméből, mig dolmányaik finomabb – vörösszinü – posztóból készültek. A tisztek ruházata kékszin selyem-posztóból vala. T. K. 110) Vagy a jászkürt (Lehel kürtje) értetik, vagy a tárogatók, a melyeket a kuruczvilág után dugdosni kellett, felsőbbségileg elrendeltetvén megégettetésök, mivel Kecskeméten egy tárogatós a „ H a j h R á k ó c z i B e r c s é n y i …“ hires kurucz tárogató- nótát elfujván rajta, a lakosságot oly tüzbe hozá, hogy ezek vasvillára, dorongra kapva, 40 vasasnémetet megöltek, s a többit elkergették. T. K. 111) Valószinüleg Prinz v. Preussen. T. K. 112) Vagy: s e b e s .
- 51. T. K. 113) Vagy: „Egészen porrá töri.“ T. K. 114) Vagy: „Ellenség elszéledvén Nemesség elterjedvén.“ 115) Saját példányomban. T. K. 116) Portára menni = portyázni; a kuruczvilágban a porta 50 egész 5–6000 emberből álló recognoscirozó csapatot is jelentett. T. K. 117) Azaz kanót. T. K. 118) Azaz: oly ringó mint a bölcső. T. K. 119) Azaz: de nem az ő m a g a v é r e . T. K. 120) Ezen versszak teljesen hiányzik Toldynál. 121) Emez viszont Szencseyben hiányzik. 122) Többnyire nők által kezelt régi hangszer: v i r g i n a l e (A „virgo“ latin szótól.) 123) Égyes: hihetően: „ é d e s “ helyett, a mi teljes értelmet ad, mig az „egyes“ itt értelmetlen volna. 124) Azaz első álmakor, mindjárt elaludta után. 125) E versszak Szencseyben hiányozván, Toldy közléséből vettem át a mű kiegészitése végett. T. K. 126) G a j d , melyből a gajdolni ige, mint dal-ból a dalolni származik, annyi mint gúnyos, csúfos, tréfás ének. Mltgos Z á d o r György hétszemélynök ur és magy. akadémiai rendes tag
- 52. birtokában létezik egy régi ének, mely e czimet viseli „ C s ú f o s g a j d “ = tréfás ének. A c s ú f s á g szó, mely a jelen sajátságosan szép és eredeti költemény második versszakában is előfordul, hajdanában egy jelentésü volt a t r é f a szóval, s itt is ezen értelemben veendő. T. K. 127) Egyed, Aegídius. T. K. 128) Zágrábi. T. K. 129) Jószágát. 130) Karó, czölöp. T. K. 131) Folyamodik; mint futamik, futamodik helyett. T. K. 132) Tréfábul. T. K. 133) Felszólalásra, felköszöntésre. T. K. 134) Titeket; Dunántúl ma is ama régi forma használtatik. T. K. 135) Bünbeesésimért, botlásaimért; e s e t : bibliai kifejezés. T. K. 136) Egyetemben, rövidítve. T. K. 137) Azaz: hamar, gyorsan, mint a sólyom röpte. T. K.
- 53. 138) Dús, gazdag; Dunántul ma is él e szóban: d u s k á s k o d n i = mindennek bővségében lenni, benne kénye-kedve szerént válogathatni. 139) Köszöntést. 140) Talán s z ő n y i ; mezőváros Komárom közelében, igy volna értelme, de ha a Szencseyben rendes dunántúli kiejtést veszszük: s z e n y i vagyis s z e n n y i : azaz s z e d n i , – igy nem lelek ezuttal értelmet. 141) Azaz: n ő ü l v e t t . T. K. 142) Azaz sok ruhával biró. 143) Illessetek. T. K. 144) A kipontozott helyek az eredetiben olvashatlanok. T. K. 145) Azaz: nagy fejü. T. K. 146) Is. 147) Pénz neme; gyra. T. K.
- 54. TARTALOM. Tájékozásul1 Vitézi és történeti énekek Hunyadi Mátyás király billikoma23 Feddő és serkentő ének26 Thúry György éneke31 Protestáns hegedős panaszló éneke34 Intő ének a magyarokhoz47 Szegedi veszedelem55 Hegedős-ének a kenyérmezei diadalról56 Protestánsok üldözéséről85 Üldözött protestánsok éneke87 Protestáns magyarok fohásza Istenhez, jó fejedelemért89 Törökök ellen hadakozó magyar vitézek fohásza94 Kátai Mihály sírfelirata97 Nagy Péter börtönéneke99 Bethlen Gábor104 Bethlen Gábor diadal-éneke110 Hungaria118 Bucquoi-ról124 Wallensteinról126 Brandenburgi Katalin keserve127 Cantio de Zólyomi135 Sárkány István halálára139 A haldokló vitéz Fodor Pál éneke144 A veszprémi hajduk levele a palotai rabló törökökhöz149 Végbéli vitézek éneke154
- 55. Tatár rabságban levő erdélyiek dala157 Régi magyar vitéz Kádárról emlékezet161 Rákóczi László börtöni éneke171 Rákóczi Lászlóról175 Rákóczi Sámuel179 Fegyvert s bátor szivet182 Balogh Zsigmond bús éneke185 Gyászének Zrínyi Miklós haláláról189 Keserv Zrínyi Miklós halálán196 Fogarasi bajnok bús éneke205 Kisfaludy Lászlóról209 Oláh Geczi213 Buga Jakab éneke218 Bujdosó éneke221 Reménység az embert225 Kovács György végbucsúdala228 Győri lutheránus templomnak elégésérül panaszolkodó ének232 Raby István éneke236 Bucsuéneke egy jegyzőnek, kinek a nánási piaczon feje vétetett 1688. sept. 20-kán242 Nagy-Kunság romlásáról249 Ideje bujdosásomnak263 Fut az oláh268 Generális insurrectio270 Nemes Jászság, hires Kunság279 Ezerhétszáz ötvenegyben283 Mohács, Mohács!286 Istenhozzád Magyarország!293 Huszártoborzó297 Bakatoborzó299 Katona bucsuja301 Megbusult katona éneke303 Páter Márton306 Megöltek egy huszárt310 Székely vitézek éneke a török hadakozáskor313
- 56. Welcome to our website – the ideal destination for book lovers and knowledge seekers. With a mission to inspire endlessly, we offer a vast collection of books, ranging from classic literary works to specialized publications, self-development books, and children's literature. Each book is a new journey of discovery, expanding knowledge and enriching the soul of the reade Our website is not just a platform for buying books, but a bridge connecting readers to the timeless values of culture and wisdom. With an elegant, user-friendly interface and an intelligent search system, we are committed to providing a quick and convenient shopping experience. Additionally, our special promotions and home delivery services ensure that you save time and fully enjoy the joy of reading. Let us accompany you on the journey of exploring knowledge and personal growth! textbookfull.com