Learning from e-Learning: Initial experiences from the European Learning Grid

1
Learning from e-Learning Initial experiences
from the European Learning Grid

• Colin Allison, Stuart Purdie, Tim Storer
• Computer Science
• University of St Andrews

2
ELeGI Project

• European Learning Grid Infrastructure is a 4 year
  EC FP6 funded Integrated Project which started
  in February 2004
• 23 partners
• 12 universities
• 2 open universities (OU and HOU)
• 4 research institutes
• 4 commercial technology companies

3
ELeGI Motivation

• dominance of the information transfer learning
  model (assumed and supported by most e-learning
  products)
• content management
• easily achieved through simple use of the Web
• lecturers job is to select, or sometimes
  produce, content
• little scope for exploratory, interactive or
  collaborative learning
• wish to foster knowledge construction, i.e.
• collaborative, experiential, personalised,
  realistic, contextualised and ubiquitous learning
  modes

4
ELeGI Approach
5
Why Grid?
Grid
Web

• Content sharing
• Presentation mark-up
• Product/download
• Producer/consumer
• bolt-on security
• Ad hoc identity management (cookies)
• Secure transactions (awkward..dont press that
  button)
• Authentication ad hoc, no cross domain

• Resource sharing
• Semantic mark-up
• Dynamic service
• Virtual communities
• detailed security model
• PKI Certificate based identity management
• Cross domain trust models
• Secure transactions

6
two aspects to the Grid

• Original Grid the anatomy of the Grid -gt
• distributed supercomputing i.e. the resources to
  be shared are high performance processing,
  networking and storage facilities
• For the former to be effective the physiology of
  the Grid had to evolve OGSA Open Grid
  Services Architecture
• secure resource sharing across multiple
  autonomous domains
• Enabling collabration across domains
• Virtual communities
• Can OGSA be used as an open framework for
  e-learning?

7
Prototypes re-engineered learning
tools/facilities/environments

• Purpose to understand the meaning of being OGSA
  conformant, and issues in implementing Grid
  learning services, without worrying about
  pedagogical goals and models
• Finesse -gt Finesse Grid Services
• BuddySpace -gt BuddySpace Grid Interface
• GRASP (OGSI.Net) -gt GRASP (WSRF.Net)
• IWT -gt IWT Grid Aware
• VCLab -gt VCLab Grid Services

8
Learning Models

• General
• Socio-constructivist, knowledge construction
• Real world , Collaborative, Experiential, Context
  aware
• Formal Learning model D18
• Focused entirely on Virtual Scientific
  Experiments
• Maps onto the IMS-Learning Design format
• Output is a sequence that feeds a virtual control
  laboratory
• personalised
• Informal Learning D20
• Social learning learning as a side effect
  collaborative environments, enhanced presence

9
Standards and Specifications

• Necessary for interoperability W3C, IETF, GGF,
  IEEE, IMS, etc.
• OASIS (Organization for the Advancement of
  Structured Information Standards)
• Web Services Resource Framework (WSRF)
• Universal Description, Discovery and Integration
  (UDDI) protocol
• Web Services Description Language (WSDL)
• Security X.509 certificates, https, ..
• Globus Toolkit 4 (GT4) open source, public
  domain, multi-platform
• WSRF.NET Microsoft Windows only

10
SEES2 EnCOrE

• Main purpose to build an informal learning
  service based on a virtual scientific community
• Collaborative knowledge capitalization
• Development of a dictionary of organic chemistry
  as a component of an electronic encyclopaedia
  EnCOrE
• Domain synthetic organic chemistry
• Informal learning
• Of the users (students, researchers, engineers)
• Of the authors about their own language and about
  concepts and techniques required to build such a
  service
• Formal learning
• Training new participants to the EnCOrE activity

11
SEES3a eQualification

• Scope Formal and Informal Training
• Testbed ASIMIL (Aeronautical Simulator)
• Training Session
• Briefing
• Simulation
• Debriefing
• The e-Qualification process
• Process allowing to qualify and certify formal
  and informal learning
• Adapted to the training domain
• Objective accreditation
• e-Qualification process is efficient
• and results are matching with the training norms

12
SEES4 Masters in ICT

• Purpose of the SEES
• Formal learning
• Remote institutions collaboration to offer same
  program
• Virtual Campus
• Common courses among remote student communities
• Virtual Classrooms
• Common scientific experiments
• Virtual Scientific Experiments
• Target group
• AIT students, CMU students, Japan CyLab students

13
SEES5 Masters in Physics at HOU

• Purpose of the SEES
• Formal learning
• Collaborative/Social Learning in Physics Course
  at HOU
• Target group
• HoU Students
• Main characteristics
• Students cooperate, perform experiments
• Knowledge construction through the exchange of
  data and knowledge
• Types of services
• Virtual Experiments
• Virtual Communities support

14
ELeGI architecture but whats really needed?
15
Core architecture
16
Current Demonstrators

• VSE demo
• Using Reload to create an IMS-LD document
• Using IWT-GA and CopperCore to control VCLab
  according to the IMS-LD script
• Generic CC and enhanced presence
• BuddySpace (deployed)
• Grid Shared Desktop (deployed)
• Interoperability of Grid services (deployed)
• Using Finesse Grid Services (FIGS) and Buddyspace
  Grid Service Interface (BUGSI) to create a
  flexible learning environment with enhanced
  presence for
• Business finance, Financial modelling
• Decision making in team management scenarios
• SensaSim Grid Service IMS-LD (piloted)

17
Evaluation Framework

• A deployment is a set of ELeGI learning events
  for an identified group of users for a specific
  period
• Deployment and Evaluation Templates
• each SEES is evaluated by multiple deployments,
  each with a separate case file
• each case file is organised into inputs and
  outputs
• Inputs include the DE document, evaluation
  materials, such as questionnaires, and results
  including completed questionnaires and system
  logs
• Outputs include analyses and summary reports

18
Case File Outline
19
Quality of Service Issues for the Grid WSRF and
GT

• ?Dangerous assumption that there will be QoS
  on demand for the Grid
• Realistic assumption Grid-based learning
  environments must make use of available
  infrastructure e.g. the Internet

20
Browser client
Web server (Tomcat)
Grid server container GT4 Java Core
Database e.g. mySQL
Simple small packet ( 100 bytes content)
http(s) tcp
jdbc-mySQL tcp
112ms
SOAP http(s)
127ms
From before tcp handshake until just before
rendering
22ms
100ms
109ms
20ms
jdbc-mySQL tcp
SOAP http(s)
http(s) tcp
21
Conclusions

• Grid offers a secure resource sharing model
• Resources are virtualised as services
• Resources/services are described using semantics
  for description and discovery
• Web is still important for delivery of services
  and user interfaces but multiple interfaces now
  possible
• Decoupling of functionality from UI has shown
  good results
• Standards and specifications helps towards
  interoperability
• Project has demonstrated interoperability using
  Grid services
• Not clear yet if the current OASIS standards give
  the all the flexibility and interoperability
  needed for e-learning framework
• Currently working on UDDI and Grid Shib

22
References

• Allison, C., S.A. Cerri, A. Gaeta, M. Gaeta, and
  P. Ritrovato, Services, Semantics and Standards
  Elements of a Learning Grid Infrastructure.
  Applied Artificial Intelligence, 2005. 19(9-10)
  p. 861-879.
• Ritrovato, P., C. Allison, S.A. Cerri, M. Gaeta,
  S. Salerno, and T. Dimitrakos, eds. Towards the
  Learning Grid, Advances in Human Learning
  Services. Frontiers in Artificial Intelligence
  and Applications, ed. J. Breuker, et al. Vol.
  127. 2005, IOS Amsterdam. 239.
• www.elegi.org

23
Thank you!

• Questions?