No Guru, No Method, No Teacher: Self-Classification and Self-Modelling of E-Learning Communities

download No Guru, No Method, No Teacher: Self-Classification and Self-Modelling of E-Learning Communities

of 14

  • date post

  • Category


  • view

  • download


Embed Size (px)


Zinayida Petrushyna, Ralf KlammaRWTH Aachen UniversityEC-TEL 2008, Maastricht, The NetherlandsSeptember 18, 2008

Transcript of No Guru, No Method, No Teacher: Self-Classification and Self-Modelling of E-Learning Communities

  • 1. Zinayida Petrushyna, Ralf Klamma RWTH Aachen University EC-TEL 2008, Maastricht, The Netherlands September 18, 2008 No Guru, No Method, No Teacher: Self-Classification and Self-Modelling of E-Learning Communities

2. Agenda

  • Motivation
  • Self-regulated life-long learning
  • Communities of Practice(CoP) model and dimensions
  • Self-monitoring of E-Learning repositories
  • Results of Self-modelling of E-Learning communities
  • Conclusions and outlook

3. Motivation

  • Firstly to learn, secondly to learn and thirdly to learn (Lenin, 1923)
  • ROADMAP(Prolearn)

V1 V2 V3 V4 V5 V6 TEPL IST challenge (3G, IpV6, nanotechnologies,convergence, web services, ambient intelligencescenario)

  • Industry challenges:
  • Performance support
  • Continuous improvement
  • Incremental development
  • Processed based integrated
  • learning)
  • Industry challenges:
  • Innovation
  • Entrepreneurship
  • ability to change
  • Competency and performance
  • management

Learners perspective: Continuous personalDevelopment Recognition and portability Of learning achievements Socio-economicSystems : Market take-up Social inclusion The Six Prolearn Vision Statements Everyone should be able to learn anything at anytime at anyplace (personalization adaptation) Learning as a means to support and enhance work performance Promote innovation and creativity and entrepreneurship Learning as a means to increase employability (flexibility and survivability of employees) Socio-economic systems market take up Access to professional learning for all extending the knowledge based society 4. Self-regulated Life-long Learning

  • Dynamic perspective on communities
  • Defining disturbances (Troll)
  • Analyzing communities
  • Application of patterns(Troll)
  • Reflecting models according
  • to the reality
  • Adapting reflected
  • models

5. CoP Model I*Modelling (Yu et. al, 1994) 6. CoP Dimensions

  • Mutual engagement(ME)
  • If you are aware of "what matters" in the scope of a community your engagement is enabled
  • Joint enterprises(JE)
  • CoPs and theirs members can follow the situations happening
  • around them and because of them
  • Shared repertoire(SR)
  • Members represent community knowledge

7. Prolearn MediaBase the Repository of Web-basedE-Learning Resources

  • Data management: Database and crawlers
  • MailWatcher refinement for Pattern Analysis
  • thread identification script
    • reply_to field
    • subject field
    • 22% of thread reduction
  • Data cleaning script for
  • thread content
    • No HTML tags and technical data
    • No duplicates (Levenshtein, 1969)

Project Project Project Mailing lists Thread Thread Thread Thread Thread Thread Thread Thread Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail Mail 8. Fundamentals andMethodologies of Monitoring

  • Network analysis and graph theories
    • G=(V, E),weightw(v), w(e) , degreed(v)
    • Centrality indices: betweenness, closeness (Wasserman & Faust, 1994)
  • Social network analysis
    • Dynamic social network analysis (Newmann et al., 2006)
    • Patterns: spammer, trolls, structural hole, innovative star, weak tie (Klamma et al., 2006)
    • Web 2.0 and social software
  • Linguistic and emotional analysis
    • Part-of-Speech tags ( Manning et al., 1999 )
    • Sentiment extraction (Mishne et al., 2006;Pennebaker et al., 2007 )

9. Mutual Engagement

  • Conectiveness, biconectiveness
    • How dense the connections?
    • How diverse the community?
  • Hubs, authorities and scale free network
    • Is it possible to differentiate between the nodes?
  • Degree centrality, closeness centrality, betweenness centrality
    • Who is central?
    • Who is the most connected?
    • Who influences mostly on the community?
  • Emotional impact
    • What categories of words is used mostly within the community?

10. Joint Enterprises &Shared Repertoire

  • Affordance
    • What are the types of processes one can execute?
    • What functions/features does the medium possess?
  • Awareness
    • Do the community members know about changes?
  • Media centric theory of learning
    • What kind of changes happens when a process is performed?
  • Sentence model + Part-of-Speech tagging
    • What is the content of the context?

11. Self-monitoring ofthe E-Learning Repository

  • Structural monitoring
    • Monologues threads
    • Reply-senders, reply-receivers
    • Communicators
    • Cross-users
  • Semantical monitoring
    • 32 word categories(4500 words)
    • psychological constructs
    • 7 personal concern
    • 3 paralinguistic dimensions

Word category Number of words in the dictionary Included words FRIENDS 36 companion, friend, mate, etc. ANGER 364 defense, rude, victim, etc. INSIGHT (understanding) 193 become,feel, inform, seem, think, etc. FILLER 8 yakno, ohwell, etc. POSEMO 405 agree, improve, support, etc. NEGEMO 495 fury, panic, temper, etc. 12. Hierarchical Clustering Results

  • Question-answer community
    • dyadic and sequent communications
    • insight and discrep words the query-explanation nature
  • Disputative community
    • reply-sender
    • explanation, disagreement and quarells the discussion nature

13. Factor Analysis Results

  • Question-answer community
    • reply sender
    • motion, social, filler, persuade and insight words

14. Conclusions & Outlook

  • E-Learning communities as CoP
  • Monitoring means (structure + semantics)
  • Modelling hypotheses with Hierarchical Clustering and Factor Analysis
  • Influence of the analysis on the learning process
  • Application of linguistic techniques on the semantic analysis
  • Methods of self-modelling of CoP
  • Recommendations in Life-long Learning based on game-theoretic models