Enabling Technologies for future learning scenarios: The Semantic Grid for Human Learning

1
Enabling Technologies for future learning
scenariosThe Semantic Grid for Human Learning
The 2nd International Workshop on Collaborative
and Learning Applications of Grid Technology and
Grid Education May 9 - 12, 2005, Cardiff, UK

• Pierluigi Ritrovato
• Research Technology Director
• Centro di Ricerca in Matematica Pura ed Applicata
• ELeGI Scientific Coordinator

2
Overview

• Background and motivations
• The ELeGI Project
• Some characteristics of future learning scenarios
• Building our vision
• The Semantic Grid for Human Learning
• Scenario
• Conclusions and future works

3
Background and motivations

• Knowledge is changing our society and our
  lifestyle
• It is the new cornerstone around which education
  (and not only education) should be re-thinked
• Information transfer based learning approaches
  are no more suitable
• Learners passivity instead of activity and
  dynamicity
• learner has no way to impact the learning process
• Too effort on defining and providing collective
  inputs (e.g. the educational contents) of the
  learning process
• No personalization, difficulties to put the
  single learner feedbacks in the learning
  process, no contextualization
• Uniformity of learning outcomes
• All have to learn everything in the same way

4
Background and motivations Current e-Learning
solutions

• Mostly e-Learning solutions are based upon the
  previous approach
• They are distance learning solutions and provide
  a digitalization of the previous approach
• e-Learning becomes an activity in which teachers
  produce, and students consume, multimedia books
  on the Web
• Missing specific didactical models
• Not any support of pedagogical aspects
• There are also some e-Learning solutions not so
  tied to the Info transfer paradigm supporting key
  aspects of the learning process
• collaboration, course personalization, virtual
  experiments
• These solutions present a common issue they are
  mainly focused only on a single aspect of the
  learning process
• What happens if my pedagogical needs change? Do I
  have to change my e-Learning solutions?

5
Background and motivationsIt is time to change

• According to us, it is time to make a process
  innovation in defining and developing e-Learning
  solutions that should support a learning process
• driven by the pedagogical needs of the learner
• in which the learner is a central and active
  figure
• in which the learning outcome (e.g. the knowledge
  creation) occurs through social interactions and
  active experiences and it is used as a feedback
  to refine the process itself

6
The ELeGI ProjectThe Project Vision

• To produce a breakthrough in current (e) Learning
  practices with the creation of a distributed and
  pervasive environment based on Grid technology
  for effective human learning where

• learning is a social activity consumed in
  communications and collaborations based dynamic
  Virtual Communities
• learners, through direct experiences, create and
  share their knowledge in a contextualised and
  personalised way

7
The ELeGI Project How we conceive the Learning

• Contextualised learning
• the understanding of concepts through direct
  experience of their manifestation in realistic
  contexts (e.g. providing access to real world
  data)
• Social learning
• the users mental processes are influenced by
  social and cultural contexts
• Collaborative learning
• more than a simple information exchange peers
  interactions, conversation tracks, knowledge
  reconstruction
• Personalised learning
• guarantee the learner to reach a cognitive
  excellence through different learning path
  tailored on learners characteristics and
  preferences

8
Some characteristics of future learning scenarios

• Distributed architecture and deployment
  environment
• Service Orientation
• teaching and training are conceived as support
  services
• The learning process is enabled or enhanced by a
  combination of services
• course material retrieving and packaging,
  tutoring, virtual meeting,
• Community, Conversation and collaboration based
• Is central for all kinds of formal and informal
  learning
• Is crucial in learning by being told but also
  in the coaching of skill acquisition
• Learner autonomy
• Freedom of decisions
• Flexibility
• Control of time, space, place, devices,
• Dynamicity
• the learner can influence the process
• The social and context aspects influence the
  learner
• High demand of interoperability
• access to Resources on heterogeneous environments
• Security and Trust
• AAA protocols, confidentiality, privacy

9
Building our vision Enabling technologies

• To build future learning scenarios we need a
  technology allowing
• autonomous and dynamic creation of communities
• active and realistic experiments
• personalization
• knowledge creation and evolution
• to reach all the features of the previous
  slide!
• Currently, we have different enabling
  technologies allowing, more or less, to create
  our vision
• Distributed Middleware, Web Services, Agent,
  Semantic Web, Grid

10
Building our vision Enabling technologies

• Distributed middleware not Service Oriented
  stable reference models for distributed
  architecture, many facilities also domain
  specific but
• Too tied to a product vision while we are closer
  to a service one
• Method based not Message based it need a lot of
  effort to implement a composition based paradigm
  useful to create personalized learning
  experiences (re-)using data, units of learning,
  knowledge and tools distributed across different
  organizations
• Web Service service based, aiming to provide
  interoperability among distributed loosely
  coupled components, good to implement a
  composition based paradigm but
• It is generally based upon a stateless model
  while the state is fundamental in conversational
  processes
• It need effort to implement resource management
  and discovery mechanisms, information and
  knowledge management, resource sharing and other
  important features of the proposed learning
  process and of a Virtual Learning Community

11
Building our vision Enabling technologies

• Agent good for personalization and
  contextualization, communities creation, goal
  oriented but
• They have to be reinforced with mechanism to
  discover, acquire, federate, and manage the
  capabilities/resources/contents needed to
  create/delivery the personalized learning
  experiences
• Semantic Web knowledge management and
  formalization, knowledge based communities and
  interactions but
• They need effort to define advanced algorithm for
  resources reservation to support efficient
  resources management allowing 3d simulations and
  immersive VR

12
Building our vision The Grid added value

• Grid technologies
• Rely upon a dynamic and stateful service model
  (e.g. WSRF or WS-I) and this affects also the
  development of learning scenarios (need for state
  in conversational processes)
• Are key technologies to build the VO paradigm (VO
  are the right place for carrying out
  collaborative learning experiences)
• As we will see later, are the most suitable to
  build IMS LD Complaint Framework (our learning
  process is pedagogical driven)
• Provide the scale of computational power and data
  storage needed to support realistic and
  experiential based learning approaches involving
  responsive resources, 3d simulations and
  immersive VR
• Are demonstrating their effectiveness for
  implementing e-Science infrastructure for sharing
  and manage data, applications and also knowledge
• Through the virtualisation and sharing of several
  kind of resources facilitate the dynamic contexts
  generation
• The dynamic service discovery and creation will
  allow the true personalisation
• Grid are becoming a glue among different
  technologies like Agent, Semantic Web, Web
  Service that, as standalone solutions, provide
  only partial benefits to our learning vision
  and our purpose is not to spend effort to fill
  the gaps

13
Building our visionIMS-LD and support for
pedagogies

• The pedagogical support is a key factor that
  distinguishes our learning approach with respect
  to other relevant learning initiative
• We need to catch all the pedagogical features
  identified and not to customize our solution for
  a single pedagogy
• IMS-LD is focused on the modeling of learning and
  teaching practices that go beyond simple
  traditional web-based LOs delivery
• the learning activities, which can be defined as
  interactions between a learner and an environment
  to achieve a planned learning outcome
• the learning approaches, involving selection and
  orchestration of the activities on the basis also
  of the pedagogies

14
Building our visionDrawbacks of IMS-LD

• In any case, IMS-LD presents some drawbacks
• LD scenarios implement domain-dependent
  pedagogies (early binding of learning objects)
• Learning processes cannot be really adaptive with
  respect to learner profiles (execution flows are
  pre-arranged at IMS-LD design-time)
• If the context (didactic domain didactic model
  learner model) of a learning scenario changes
  I need new contents and services suitable for the
  new context and I have to bind them statically!
• LD scenarios dont exploit the advantages of a
  dynamic distributed environment in which I can
  dynamically find and bind new contents and
  services

15
Building our visionImproving IMS-LD for our needs

• Our solution to the previous drawbacks is to
  provide
• extensions for IMS-LD in order to define
  domain-independent pedagogies
• a knowledge model able to describe educational
  domains and learners using respectively
  ontologies and learner profiles
• a set of algorithms for automatic building of
  personalized learning paths pulled out from
  ontologies using learner profiles and target
  concepts
• a set of algorithms able to join personalized
  learning paths with domain-independent pedagogies
  obtaining an abstract unit of learning (no
  binding with real learning objects) where each
  concept in learning path is explained using the
  selected pedagogy
• an abstract unit of learning run-time model
  interacting with a Grid system able to provide
  (at run-time) a late binding with desired
  learning objects and services

16
Building our visionThe Grid added value to
learning scenarios

• Grid technologies provide advanced mechanisms for
  automatic discovery and binding of new suitable
  contents and services as well as self-adaptive
  mechanisms when deploying the LD scenarios and,
  obviously, the learning activities composing a
  scenario
• Grid provides dynamicity and adaptiveness to LD
  scenarios

17
The Semantic Grid for Human Learning

• The Semantic Grid for Human Learning can be
  defined as a domain verticalization of the
  Semantic Grid improved with tools, services,
  languages, standards and technologies for the
  Education
• WSRFWSRP for enhancing the underlying service
  model (dynamic, stateful and presentation-oriented
  )
• Semantically enriched services typical of the
  learning domain
• IWT Grid-Aware Base Services providing
  functionalities typical of a Learning Management
  System
• IWT Grid-Aware Learning Services providing
  high-level functionalities for a personalized
  learning experience
• Driver Service WSRP compliant services providing
  the full management (creation, delivery, update)
  of a didactical resource
• IMS-LD for creating learning scenarios able to
  catch all the identified pedagogical features
• User Centric Portal implementing the behaviour of
  a WSRP consumer
• allowing an easy customization and administration
  of community portals
• And, obviously, the standards, specifications and
  technologies providing the foundation of the
  Semantic Grid (e.g. Data Services, OWL, OWL-S, )

18
The Semantic Grid for Human Learning
This is a work in progress
19
ScenarioExtending IMS-LD with IMS-MD attributes

• A simple inductive didactic method

20
ScenarioAbstract Unit of LearningPlunging
didactic methods into didactic domains

• Explanation of some Calculus concepts by
  inductive method

21
ScenarioRunning unit of learning
22
Conclusion and future worksAssessment with other
initiative

• OKI open specifications that describe how the
  components of a learning environment communicate
  with each other and with other campus systems
• Sakai Collaboration and Learning Environment by
  exploiting the OSIDs defined in the frame of OKI
  and Grid Service-Oriented portals (OGCE)
• Commonalities between Sakai and our solution
• service concept and SOA, Grid Service Oriented
  portal based on the portlet concept, some lowest
  level and higher level OSIDs find a mirror in our
  IWT Grid-Aware Base and Learning Services and
  other OSIDs overlap with some Grid standards and
  specification (SQL lt -- gt OGSA DAI)
• Differences between Sakai and our solution
• We aims to support pedagogies, we are less
  content oriented, and our solution is more
  focalized on knowledge management and
  collaboration through social interactions (not
  only collaborative tool)

23
Conclusion and future worksAssessment with other
initiative

• JISC ELF it is part of a wider e-Learning
  programme focused on four themes e-learning and
  pedagogy technical framework and tools for
  e-learning innovation and distributed e-learning
• ELeGI project is very close to the e-Learning
  programme of the JISC (We have the focus on the
  same themes)
• We have identified a technology and also a set of
  services specific for a VLC while ELF is yet
  neutral form these viewpoints

24
Conclusion and future works

• Clear benefits for educational community can come
  from a well defined Grid based strategy and Grid
  community must start to fill the gaps among
  powerful general visions (like the Semantic Grid)
  and practical requirements of many e-Research and
  e-Business communities
• We have discussed of how to customize the
  Semantic Grid vision for the Education Training
  and we hope that similar efforts related to other
  fields of e-Research may arise
• Very next steps
• To complete the development of IWT Grid-Aware
• To refine the Semantic Grid for Human Learning
  Architecture
• To define more complex future learning scenarios

25
Thank you very much for your attention
Contacts ritrovato_at_crmpa.unisa.it