2015 03 19 (EDUCON2015) eMadrid UPM Towards a Learning Analytics Approach for Supporting discovery and reuse of OER. An approach based on Social Networks Analysis and Linked Open Data

Towards a Learning Analytics Approach for
Supporting discovery and reuse of OER
An approach based on Social Networks Analysis and
Linked Open Data
EDUCON 2015 18-20 March 2015,
Tallinn University of Technology, Tallinn, Estonia
Edmundo Tovar Caro (presenter)
Universidad Politécnica de Madrid
edmundo.tovar@upm.es
Nelson Piedra, Janneth Chicaiza, Jorge López
Universidad Técnica Particular de Loja, Ecuador
{nopiedra@utpl.edu.ec, jachicaiza, jalopez2}@utpl.edu.ec

A traditional understanding
of Learning Analytics
• The most common type of learning analytics is based on the
analysis of extracted learning data from online courses with
data mining and other data processing techniques, often
followed by some form of reporting and visualization of data.
• Learning analytics help in identifying the effectiveness of
learning elements:
• can help in engaging students
• can guide teachers in the preparation and deployment of the
teaching activity.
• Learning Analytics seeks to produce ‘actionable intelligence’;
the key is that action is taken. Five steps: Capture, Report,
Predict, Act, Refine.

• Learning Analytics (LA) is an emerging domain for
educational research and IT-supported learning
processes. The concept is broad and applies to
different approaches to the use of educational data
oriented toward decision-making and personalization
of learning environments.

OER Social Motivation
• The right to education is part of the economic, social and cultural
rights defined in the Universal Declaration of Human Rights.
Education shall be free.
• The main purpose of OER movement is to provide open and free
access to high quality digital learning materials. For this reason, its
effects or implications on higher education are well known.
• OERs are distributed across a variety of repositories, using a variety
of metadata standards, various interfaces and access mechanism.
• Harvesting OER for use elsewhere can involve complicated
metadata crosswalk, and metadata often is incomplete and out of
date.

The OERs movement
• MIT OpenCourseWare (OCW) in 2001provided free, virtual
and non-commercial access to MIT courses to educators,
students and self-learners around the world
• The next aim was to create a flexible movement based on
an efficient model that other universities could emulate at
the moment of publishing their own educational materials,
generating synergies and spaces of collaboration.

OER Repositories, current status:
A world of OER Silos
A New Silo: A New Obstacle
Image Source: http://www.planetdevops.net/?cat=181

OER challenges inherent to
discovering and reuse
• The OER movement poses challenges inherent
to discovering and reuse digital educational
materials from highly heterogeneous and
distributed digital repositories.
• To find OER on the Web today, users must first
be well informed of which OER repositories
potentially contain the data they want and what
data model describes these datasets, before
using this information to create structured
queries.
Elena Berriolo, Drawing for The Silo, 2010; courtesy Raphael Rubinstein

Our Purpose
• A learning analytics approach for learning OERs unifying
two complementary perspectives:
• (1) structural analysis of the OER content to integrate and
interoperate OER materials;
• (2) content analysis to capture dynamics in OER content using
SNA. The information that can answer a query to discover OERs
should be already available on the Web as linked data.
This approach allows the cross-cultural, cross OER
providers, and cross-subjects comparisons in large scale.

2015 03 19 (EDUCON2015) eMadrid UPM Towards a Learning Analytics Approach for Supporting discovery and reuse of OER. An approach based on Social Networks Analysis and Linked Open Data

In previos works, we apply the Linked Data Design Issues to
explore, visualize and use information that is semantically
related to open educational resources that are accessible via
the OCW Consortium. Serendipity initiative has begun to adopt
models that are suitable for expressing OCW and OER
metadata as Linked Data.
OCW resource metadata can be enriched using datasets
hosted by the Linked Open Data cloud. Additionally, the Linked
OER and OCW Data environment enabled us to discover and
reuse open educational materials.
http://serendipity.utpl.edu.ec
http://serendipity.utpl.edu.ec/map

A World of Silos
(individual datasets and
specialized research tools)
A graph-data model enables users to find hidden connection that span across many repositories,
revealing new insights into biology, enabling users to ask complex questions. Users can look for
relationships in the data.
A World of Linked Data
(Integration way standard taxonomies,
ontologies, vocabularies, and
structured data repositories)
Workflow in which existing
datasets are transformed,
integrated and aggregated
into a normalize data
collection to interoperate,
and integrate pieces of data
from existing datasets
Apps

A comprehensive learning analytics for Open Educational Resources based on Linked Data
design issues and Social Network Analysis approach
Linked data exposes previously siloed repositories as Web-data
graphs, which can be interlinked, integrated and interoperated with
other datasets, creating a global-scale interlinked data space.
Educon2015-@nopiedra

Linked Data is the way that the Semantic Web has to link and interoperate data that are distributed on the Web,
so that they are referenced in the same way they do the links of the web pages. Furthermore, connect, query and
recombine data from the Web, as if they were simply part of a global database. These advances can be a way to
support interoperability, accessibility and reusability of all types of data.

Social Networks Analysis to understand and
visualize the patterning of interactions
betweem individuals in a quantitative manner.
• Social network analysis (SNA) is a framework used to study the
structure of social networks.
• The power of social networks analysis it is to assume that the attributes of
individual nodes are less important than their relationships and ties with
other nodes within the network.
• This approach has turned out to be useful for explaining many real-
world situations. Social networks analysis have also been used to
examine how nodes interact with each other.

Centrality, a structural
characteristic of nodes in a graph.
• Centrality, is a structural characteristic of individuals in the
social network. Centrality refers to indicators, which identify
the most important vertices within a graph.
• A centrality score means that individual fits within the
network overall.

Types of centrality
• Closeness centrality, this measure expresses the average social
distance from each individual to every other individual in the
network.
• Betweenness centrality is another measure that is derived from the
concept of counting the shortest paths between individuals in a
network. This measure helps to find the individuals who are
necessary conduits for information that must traverse disparate parts
of the network.
• Eigenvector centrality, this measure basically indicates the extent to
which an individual is a node with many connections connected with
others high-profile nodes. Individuals with high eigenvector scores
have many connections, and are leaders of the network, and their
connections have many connections, out to the end of the network.

Visualization Data
• Users should explore datasets even if the publisher of the
data does not provide any exploration or visualization
means.
• Applying information visualization techniques to the OER
data helps users to explore large amounts of data and
interact with them.

Geo Data Visualization of OERs and relationated subjects

Applications: Feedback,
adaptation and personalization
• OER creators questions:
• Which ins the importance of knowledge branches, networks?
• OER providers questions:
• which is the utility of their resources,
• who reuses and adapts them.
• Users could have answers to questions such as:
• What actions should be suggested for the user?
• What groups do users cluster into?
• How should the user experience about discovery or reuse be changed for the next user?

Conclusions
• This is the vision of an OER learning analytics approach, which can be used by
all stakeholders involved in discovery and reuse of open educational resources.
• The availability of Web-scale information in a structured and fine-grained
representation could generate a paradigmatic shift in how applications and
users consume and share data. Linked data provides a data layer on the Web
that represents objects and relations.
• The datasets information can be used to support different aspects of the open
educational process; linked data within this schema can be analyzed and mined,
and also used for formulating recommendations to the users.

Thanks!!
EDUCON 2015 18-20 March 2015,
Tallinn University of Technology, Tallinn, Estonia
Towards a Learning Analytics Approach for
Supporting discovery and reuse of OER
An approach based on Social Networks Analysis and
Linked Open Data

2015 03 19 (EDUCON2015) eMadrid UPM Towards a Learning Analytics Approach for Supporting discovery and reuse of OER. An approach based on Social Networks Analysis and Linked Open Data

More Related Content

What's hot (19)

Similar to 2015 03 19 (EDUCON2015) eMadrid UPM Towards a Learning Analytics Approach for Supporting discovery and reuse of OER. An approach based on Social Networks Analysis and Linked Open Data (20)

More from eMadrid network (20)

Recently uploaded (20)

2015 03 19 (EDUCON2015) eMadrid UPM Towards a Learning Analytics Approach for Supporting discovery and reuse of OER. An approach based on Social Networks Analysis and Linked Open Data

Editor's Notes