Use of Contextualized Attention Metadata for Ranking and Recommending Learning Objects Xavier Ochoa, ESPOL, Ecuador Erik D...
Agenda <ul><li>What is the problem? </li></ul><ul><li>Use of CAM for Ranking and Recommending </li></ul><ul><li>Demo: Shar...
Human Review
Human Review
TFIDF
What is the problem? <ul><li>A Ranking mechanism must be:  </li></ul><ul><li>MEANINGFUL and SCALABLE </li></ul>
Ranking Strategies for LO <ul><li>There is a lot of ideas that could help to implement a Ranking Mechanism for Learning Ob...
Network Analysis <ul><li>There is not a linking structure between LO like in Web pages or Scientific papers </li></ul><ul>...
Lifecycle  Actions Main Information Source Creation Creating author, components Authoring tool, Components Labeling Labeli...
Network Analysis <ul><li>CAM as K-Partite Graph </li></ul>O 1 O 2 O 3 C 1 C 2 U 1 U 2 A 1 A 2 User Partition Course Partit...
Network Analysis <ul><li>CAM as 2-Partite Graph (User-Object) </li></ul>
Network Analysis Metrics <ul><li>From this Graph we can create serveral metrics. For example: </li></ul><ul><ul><li>Popula...
Similarity Metric <ul><li>The K-Partites Graphs generated from CAM can be folded into Normal Graphs </li></ul><ul><li>Thes...
Similarity Metric
Similarity Metric <ul><li>Object Similarity based on Downloads </li></ul><ul><li>Object Similarity based on Re-Use </li></...
Object Similarity
Communities ARIADNE
Personalized Rank <ul><li>We can create a profile of the user based on its CAM </li></ul><ul><li>We will use the same LOM ...
Contextual Recommending <ul><li>If the CAM is considered not only as a source for historic data, but also as a continuous ...
Demo
Conclusions <ul><li>Meaningful and Scalable Ranking Metrics can be created for Learning Objects as has been done for other...
Thank You! <ul><li>Questions, Comments, Critics….   </li></ul><ul><ul><ul><li>Are all welcome!! </li></ul></ul></ul>
Upcoming SlideShare
Loading in …5
×

Use of Contextualized Attention Metadata for Ranking and Recommending Learning Objects

1,392 views

Published on

Presentation at CIKM 2006 about how to calculate the relevance of a learning object based on the attention that the object have received

Published in: Technology
  • Be the first to comment

Use of Contextualized Attention Metadata for Ranking and Recommending Learning Objects

  1. 1. Use of Contextualized Attention Metadata for Ranking and Recommending Learning Objects Xavier Ochoa, ESPOL, Ecuador Erik Duval, KULeuven, Belgium
  2. 2. Agenda <ul><li>What is the problem? </li></ul><ul><li>Use of CAM for Ranking and Recommending </li></ul><ul><li>Demo: ShareLOk </li></ul><ul><li>Conclusions </li></ul>
  3. 3. Human Review
  4. 4. Human Review
  5. 5. TFIDF
  6. 6. What is the problem? <ul><li>A Ranking mechanism must be: </li></ul><ul><li>MEANINGFUL and SCALABLE </li></ul>
  7. 7. Ranking Strategies for LO <ul><li>There is a lot of ideas that could help to implement a Ranking Mechanism for Learning Objects </li></ul><ul><li>BUT THE SEMANTICS ARE DIFERENT </li></ul><ul><ul><li>Web pages  Papers  LO </li></ul></ul><ul><li>Eg. A link from one page to other means support. A link from one LO to other does not carry the same meaning. </li></ul>
  8. 8. Network Analysis <ul><li>There is not a linking structure between LO like in Web pages or Scientific papers </li></ul><ul><li>But we have plenty of relations between different entities and LO </li></ul><ul><ul><li>Authors, Instructors, Readers, Courses, LMS </li></ul></ul><ul><li>The main source of this information can be extracted from Contextualized Attention Metadata (CAM) </li></ul>
  9. 9. Lifecycle Actions Main Information Source Creation Creating author, components Authoring tool, Components Labeling Labeling metadata format, origin, confidence Authoring tool or Metadata generator Offering Inserting inserter LOR or Sharing app. Selecting Searching query, results LOR’s search tool Recommending objects recommended Recommender Browsing Time LOR or Recommender Selecting object identifier LOR or Recommender Using Publicating LMS context LMS Sequencing list of sequenced objects ID tool or Packager Viewing Time, tool used Browser or app. Annotating rate or review LMS Retaining Retaining decision to keep or delete LMS
  10. 10. Network Analysis <ul><li>CAM as K-Partite Graph </li></ul>O 1 O 2 O 3 C 1 C 2 U 1 U 2 A 1 A 2 User Partition Course Partition Author Partition Object Partition
  11. 11. Network Analysis <ul><li>CAM as 2-Partite Graph (User-Object) </li></ul>
  12. 12. Network Analysis Metrics <ul><li>From this Graph we can create serveral metrics. For example: </li></ul><ul><ul><li>Popularity Rank (PR) </li></ul></ul><ul><ul><li>Author-Corrected Popularity Rank (ACPR) </li></ul></ul><ul><ul><li>Weighted Popularity Rank (WPR) </li></ul></ul><ul><ul><li>“ Best-Sellers” Rank (BSR) </li></ul></ul>
  13. 13. Similarity Metric <ul><li>The K-Partites Graphs generated from CAM can be folded into Normal Graphs </li></ul><ul><li>These graphs have a link between two nodes in the same partition, if they have been linked to the same object in the folded partition. </li></ul>
  14. 14. Similarity Metric
  15. 15. Similarity Metric <ul><li>Object Similarity based on Downloads </li></ul><ul><li>Object Similarity based on Re-Use </li></ul><ul><li>Users similarity based on Downloads </li></ul><ul><li>Author similarity based on Re-Use of Components </li></ul>
  16. 16. Object Similarity
  17. 17. Communities ARIADNE
  18. 18. Personalized Rank <ul><li>We can create a profile of the user based on its CAM </li></ul><ul><li>We will use the same LOM record to store this profile </li></ul><ul><li>Instead of having a crisp preference for a value, the user will have a fuzzy set with different degrees of “preference” for all the possible values. </li></ul>
  19. 19. Contextual Recommending <ul><li>If the CAM is considered not only as a source for historic data, but also as a continuous stream of contextualized attention information, we can use very recent CAM to generate recommendations based on what the user is focusing his/her attention at the moment. </li></ul>
  20. 20. Demo
  21. 21. Conclusions <ul><li>Meaningful and Scalable Ranking Metrics can be created for Learning Objects as has been done for other fields </li></ul><ul><li>But first we need to measure, analyze and understand the data about Learning Objects… We need Learnometrics </li></ul><ul><li>Development in Learnometrics could be easily translated in better tools </li></ul>
  22. 22. Thank You! <ul><li>Questions, Comments, Critics…. </li></ul><ul><ul><ul><li>Are all welcome!! </li></ul></ul></ul>

×