Harnessing Collective Intelligence in Personal Learning Environments
2 likes577 views
The document discusses harnessing collective intelligence in personal learning environments (PLEs). It describes how PLEs aim to put learners at the center of lifelong and informal learning through self-defined collections of tools and services. However, the abundance of information available creates knowledge overload. The document proposes a social filtering approach called a Personal Learning Environment Management (PLEM) service that harnesses collective intelligence to help learners locate quality learning resources based on social interactions with those resources. An evaluation of PLEM found average user satisfaction of 66 out of 100 but improvements are still needed in learnability and usability. Future work will explore different recommendation techniques like collaborative and content-based filtering to better adapt algorithms used in PLEM.
![Knowledge Filters
Knowledge Filters
Personal Network
Recommender Systems
The Wisdom of Crowds / Collective Intelligence
None of us is smarter than all of us [Surowiecki, 2004]
7](https://image.slidesharecdn.com/icalt2012chatti-120816070803-phpapp01/85/Harnessing-Collective-Intelligence-in-Personal-Learning-Environments-7-320.jpg)
Harnessing Collective Intelligence in Personal Learning Environments
- 1. Harnessing Collective Intelligence in Personal Learning Environments Mohamed Amine Chatti, Ulrik Schroeder, Hendrik Thüs, Simona Dakova Informatik 9 (Learning Technologies), RWTH Aachen University 1
- 2. Overview Personal Learning Environments (PLEs) Knowledge Overload Social Filtering A Service for Personal Learning Management (PLEM) 2
- 3. Personal Learning Environments (PLE) Pedagogical Perspective The environment in which I learn A more natural and learner‐centric Lifelong Learning Informal Learning model to learning Put the learner at the center Personal Learning Environments Network Learning PLE: Convergence of lifelong, Self‐Organized Learning informal, network, and personalized learning 3
- 4. Personal Learning Environments (PLE) Technical Perspective PLE: Self‐defined collection of services, tools and devices that help learners build their PKNs and learn Personal Knowledge Network (PKN): • Tacit Knowledge Nodes (People) • Explicit Knowledge Nodes (Information)
- 5. LMS vs. PLE LMS PLE Content-centric Learner-centric Management Sharing Pre-defined selection of tools Learner needs first, tool selection second One-size-fits-all Personal, responsive Formal learning Support Informal and lifelong learning support Centralized, closed, bounded Distributed, loosely coupled, open Structured, heavyweight, rigid Freeform, lightweight, flexible Top-down, hierarchical Bottom-up, emergent Command&control, one-way flow of Symmetric relationships knowledge Knowledge-push Knowledge-pull 5
- 6. From Scarcity to Abundance PLE: From knowledge‐push to knowledge‐pull Abundant access to information Knowledge Overload Need for knowledge filters 6
- 7. Knowledge Filters Knowledge Filters Personal Network Recommender Systems The Wisdom of Crowds / Collective Intelligence None of us is smarter than all of us [Surowiecki, 2004] 7
- 8. PLEM PLEM: A Social Software for Personal LEarning Management Goal: Harnessing collective intelligence to locate quality knowledge nodes (learning resources, services, experts) Social interaction metrics (e.g. Facebook, Twitter, Digg, Delicious) 8
- 9. Ranking in PLEM Ranking of learning elements based on social interaction metrics Idea: Consider each simple interaction with a learning element as a vote The learning element that gets the most votes goes first on the list subprogra.informatik.rwth-aachen.de:8180/PLEM/ 9
- 10. User Evaluation Online Questionnaire 22 evaluators Usability Evaluation Subset of the 50‐question database of the Software Usability Measurement Inventory (SUMI) Evaluation using the System Usability Scale (SUS) Average user satisfaction of 66 points out of 100 points Needs improvements in terms of system learnability and user interface Functionality Questions (ranking quality) 10
- 11. Future Work: Recommendation Techniques Examples Memory‐based algorithms: Neighborhood‐based CF Collaborative Top‐N recommendation Model‐based algorithms: Machine learning / data mining algorithms Content based Information retrieval Hybrid Combination of collaborative and content‐based approaches 11
- 12. Recommendation in PLEM Memory-based tag-based CF K Nearest neighbour Analysis Offline and Evaluation adaptation Model-based User of tag-based CF Evaluation algorithms Dimension reduction Classification Clustering Association 12
- 13. Thank You [email protected] mohamedaminechatti.blogspot.com 13