Modeling the Macro-Behavior of Learning Object Repositories
Download as PPTX, PDF1 like318 views
The document discusses the modeling of macro-behavior in learning object repositories, emphasizing how the simple process of publishing objects leads to complex interactions among contributors and consumers. It highlights the importance of engagement and provides a framework for modeling contributor growth, publication rates, and repository validation through simulations. The conclusions suggest that while the model has limitations, it allows for future testing and improvement of repository growth models.
Modeling the Macro-Behavior of Learning Object Repositories
- 1. Modeling the Macro-Behavior of Learning Object RepositoriesXavier OchoaEscuela Superior Politécnica del Litoral
- 3. Publishing Learning ObjectsIt is a “simple” process:Upload or point to the materialFill some metadataShare!
- 4. Publishing Learning ObjectsThis simple process determines the micro-behavior of contributors and consumersThis give rise to complex macro-behavior at the repository level once hundreds or thousands of individuals are aggregated
- 5. From Micro to MacroStudied for other fieldsPublication of papersApplication for patentsEconomic transactions
- 6. Growth in ObjectsSome grow linearly others exponentially
- 7. Objects per ContributorHeavy-tailed distributions (no bell curve)LORP - LORFLotka “fat-tail”
- 8. Objects per ContributorHeavy-tailed distributions (no bell curve)OCW - LMSWeibull “fat-belly”
- 9. Objects per ContributorHeavy-tailed distributions (no bell curve)IRExtreme Lotka“big-head”
- 10. Objects per Contributor – Impl.There is no such thing as an “average user”
- 11. Engagement is the key
- 12. Enagement is the keyLMSs are the best type of Repository!!!
- 13. Modeling LORPublication Rate Distribution (PRD)Lifetime Distribution (LTD)Contributor Growth Function (CGF)
- 14. Modeling LORThe period of time, measured in days is selected.The Contributor Growth Function (CGF) is used to calculate the size of the contributor population A virtual population of contributors of the calculated size is created.For each contributor: the two basic characteristics, publication rate and lifetime are assigned (PRD) and (LTD)Each contributor is assigned a starting date (CGF). The simulation is run
- 15. Modeling LOR
- 16. Model ValidationTo validate this model we compare the simulated results against the data extracted from real repositories. Three characteristics of the repository are compared: distribution of the number of publications among contributors (N)the shape of the content growth function (GF)the final size of the repository (S).
- 17. Model Validation Parameter Estimation
- 18. Model ValidationComparison of results N
- 19. Model Validation
- 24. ConclusionsSimple assumptions:how frequently the contributors publish material (publication rate)how much time they persist in their publication efforts (lifetime)at which rate they arrive at the repository (contributor growth function). Predict:distribution of publications among contributorsthe shape of the content growth functionfinal size of the repository.
- 25. ConclusionsSimple model that presents errors… but it is TESTABLENew models can be constructed and tested to determine if they are better or worstGive a way to measure the goodness of the ideas
- 26. ConclusionsAltering the lifetime distribution (that is engagement) change the kind of growth of the repository
- 27. Gracias / Obrigado / Thank youXavier Ochoaxavier@cti.espol.edu.echttp://ariadne.cti.espol.edu.ec/xavierTwitter: @xaoch