Personalized Recommender Systems for               Resource-Based Learning                 Ranking Learning Resources in F...
Resource-Based Learning                          KOM – Multimedia Communications Lab   2
Challenge: What is relevant to me right now?                                        KOM – Multimedia Communications Lab   3
Solution: Ranking of Learning Resources                                          KOM – Multimedia Communications Lab   4
Overview1. Basics2. Application Scenario: CROKODIL3. Related Work    Ranking Algorithms in Folksonomies    Recommender S...
Folksonomy SystemsA folksonomy is a system of classification derived from the practice of  collaboratively creating and ma...
Folksonomy ModelA folksonomy is a quadrupleF:= (U, T, R, Y)whereU - UsersT - TagsR - ResourcesY ⊆ R × T × U - tag assignme...
Overview1. Basics2. Application Scenario: CROKODIL3. Related Work    Ranking Algorithms in Folksonomies    Recommender S...
Application Scenario: CROKODILCROKODIL is a platform offering support for resource-based learning in professional educatio...
CROKODIL Extends the Folksonomy Model …                                   KOM – Multimedia Communications Lab 10
Semantic Tag Types                     KOM – Multimedia Communications Lab 11
Learner Groups and Friendships                                 KOM – Multimedia Communications Lab 12
Activities             KOM – Multimedia Communications Lab 13
Overview1. Basics2. Application Scenario: CROKODIL3. Related Work    Ranking Algorithms in Folksonomies    Recommender S...
Recommendation TechniquesCollaborative Filtering Recommendation Techniques:  Nearest Neighbor (cosine, correlation)  Clu...
PageRank Algorithm (Page & Brin, 1998)                                                  KOM – Multimedia Communications La...
Ranking Algorithms in Folksonomies                                     KOM – Multimedia Communications Lab 17
Ranking Algorithms in Folksonomies Ranking Strategy     Applicable For        Topic-sensitive        Group-sensitive      ...
Recommender Systems in E-LearningRecommender Systems                       DescriptionsReMashed                           ...
Overview1. Basics2. Application Scenario: CROKODIL3. Related Work    Ranking Algorithms in Folksonomies    Recommender S...
Research QuestionWhat semantic information in folksonomies can be exploited to rank learningresources in graph-based recom...
Objectives1. Investigate ranking algorithms and graph-based recommender   techniques for folksonomies2. Design and impleme...
Research Approach                    KOM – Multimedia Communications Lab 23
Current Progress: Conceptual Architecture[Anjorin et al, ISWC Workshop 2011]                                        KOM – ...
Current Progress: 3a & 3b      3a            3b      3c                    3d                                 KOM – Multim...
Tag Weights based on Semantic Tag Types Tag weights are determined based on the usage frequency of  semantic tag types   ...
Graph-Based Recommendations using Semantic Tag Types                        0.35 - 0.20 = 0.15                            ...
Graph-Based Recommendations usingSemantic Tag TypesApproach:   Traverse the links between activities to find relevant res...
Next StepsAnalyze Ranking Algorithms for FolksonomiesInvestigate Semantic Information in the application scenarioCROKODILI...
Overview1. Basics2. Application Scenario: CROKODIL3. Related Work    Ranking Algorithms in Folksonomies    Recommender S...
Future Work: 3c & 3d      3a           3b   3c                    3d                             KOM – Multimedia Communic...
Future Work: Recommendation Feedback Loop Recommendation Feedback Loop    Rank Resources    Explain recommendations    ...
Future Work: Evaluations 5 & 6                                 KOM – Multimedia Communications Lab 33
Overview1. Basics2. Application Scenario: CROKODIL3. Related Work    Ranking Algorithms in Folksonomies    Recommender S...
Personalized Recommender Systems forResource-Based LearningMotivation  Due to the vast amount of resources               ...
Publications[ARS11]    Mojisola Anjorin, Christoph Rensing, Ralf Steinmetz: Towards Ranking in Folksonomies for           ...
Literature[ Abel et al, 2008]        Fabian Abel, Nicola Henze, and Daniel Krause. Analyzing                           Ran...
Any Questions ?                  KOM – Multimedia Communications Lab 38
Upcoming SlideShare
Loading in...5
×

Erster f vortrag_personalized_rec_sys_for_rbl__20110919_ma_v5.0

823

Published on

Personalized Recommender Systems for Resource-Based Learning

Published in: Technology, Education
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
823
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
6
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Erster f vortrag_personalized_rec_sys_for_rbl__20110919_ma_v5.0

  1. 1. Personalized Recommender Systems for Resource-Based Learning Ranking Learning Resources in Folksonomies httc – Hessian Telemedia Technology Competence-Center e.V - www.httc.de KOM - Multimedia Communications Lab Prof. Dr.-Ing. Ralf Steinmetz (Director) Dept. of Electrical Engineering and Information Technology Dept. of Computer Science (adjunct Professor)Dipl. –Inform. Mojisola Anjorin TUD – Technische Universität Darmstadt Rundeturmstr. 10, D-64283 Darmstadt, GermanyMojisola.Anjorin@KOM.tu-darmstadt.de Tel.+49 6151 166150, Fax. +49 6151 166152Tel.+49 6151 166160 www.KOM.tu-darmstadt.deErster_F_Vortrag_Personalized_Rec_Sys_for_RBL__20110919_MA_v5.0.ppt 2. November 2011© 2011 author(s) of these slides including research results from the KOM research network and TU Darmstadt. Otherwise it is specified at the respective slide
  2. 2. Resource-Based Learning KOM – Multimedia Communications Lab 2
  3. 3. Challenge: What is relevant to me right now? KOM – Multimedia Communications Lab 3
  4. 4. Solution: Ranking of Learning Resources KOM – Multimedia Communications Lab 4
  5. 5. Overview1. Basics2. Application Scenario: CROKODIL3. Related Work  Ranking Algorithms in Folksonomies  Recommender Systems in E-learning4. Research Topic  Research Question  Objectives  Research Approach  Current Progress5. Future Work6. Summary KOM – Multimedia Communications Lab 5
  6. 6. Folksonomy SystemsA folksonomy is a system of classification derived from the practice of collaboratively creating and managing tags to annotate and categorize content. [Peters, 2009]a.k.a Social Tagging Systems, Collaborative Tagging Systems KOM – Multimedia Communications Lab 6
  7. 7. Folksonomy ModelA folksonomy is a quadrupleF:= (U, T, R, Y)whereU - UsersT - TagsR - ResourcesY ⊆ R × T × U - tag assignment[Hotho et al, 2006] KOM – Multimedia Communications Lab 7
  8. 8. Overview1. Basics2. Application Scenario: CROKODIL3. Related Work  Ranking Algorithms in Folksonomies  Recommender Systems in E-learning4. Research Topic  Research Question  Objectives  Research Approach  Current Progress5. Future Work6. Summary KOM – Multimedia Communications Lab 8
  9. 9. Application Scenario: CROKODILCROKODIL is a platform offering support for resource-based learning in professional education  Semantic Tag Types  Activities  Learner Groups and Friendships  Recommendations [Anjorin et al, EC-TEL 2011] KOM – Multimedia Communications Lab 9
  10. 10. CROKODIL Extends the Folksonomy Model … KOM – Multimedia Communications Lab 10
  11. 11. Semantic Tag Types KOM – Multimedia Communications Lab 11
  12. 12. Learner Groups and Friendships KOM – Multimedia Communications Lab 12
  13. 13. Activities KOM – Multimedia Communications Lab 13
  14. 14. Overview1. Basics2. Application Scenario: CROKODIL3. Related Work  Ranking Algorithms in Folksonomies  Recommender Systems in E-learning4. Research Topic  Research Question  Objectives  Research Approach  Current Progress5. Future Work6. Summary KOM – Multimedia Communications Lab 14
  15. 15. Recommendation TechniquesCollaborative Filtering Recommendation Techniques:  Nearest Neighbor (cosine, correlation)  Clustering  Bayesian Networks  Neural Networks  Probabilistic Models  Graph-Based Techniques KOM – Multimedia Communications Lab 15
  16. 16. PageRank Algorithm (Page & Brin, 1998) KOM – Multimedia Communications Lab 16 Wikipedia Commons. An art draw drawn by Felipe Micaroni Lalli
  17. 17. Ranking Algorithms in Folksonomies KOM – Multimedia Communications Lab 17
  18. 18. Ranking Algorithms in Folksonomies Ranking Strategy Applicable For Topic-sensitive Group-sensitive (adapts to context) FolkRank Users, Tags, Resources Yes No GFolkRank Users, Tags, Resources Yes Yes GFolkRank+ Users, Tags, Resources Yes Yes GRank Resources Yes Yes SocialPageRank Resources No No Personalized Resources Yes No SocialPageRank [Abel et al, 2008] KOM – Multimedia Communications Lab 18
  19. 19. Recommender Systems in E-LearningRecommender Systems DescriptionsReMashed  Recommendations for Web 2.0 content[Drachsler et al. 2009]  User-based collaborative filtering  Informal Learning NetworksRACOFI Recommendations of audio Learning(Rule-Applying Collaborative Filtering) Objects[Anderson et al. 2003; Lemire 2005] Rule-based and Collaborative filtering Using domain taxonomiesRPL recommender  Hybrid recommender system[Khribi et al. 2009]  Rated recommendations  Learning at work for specific tasks[Manouselis et al, 2011] KOM – Multimedia Communications Lab 19
  20. 20. Overview1. Basics2. Application Scenario: CROKODIL3. Related Work  Ranking Algorithms in Folksonomies  Recommender Systems in E-learning4. Research Topic  Research Question  Objectives  Research Approach  Current Progress5. Future Work6. Summary KOM – Multimedia Communications Lab 20
  21. 21. Research QuestionWhat semantic information in folksonomies can be exploited to rank learningresources in graph-based recommender systems?How can these be used to provide personalized recommendations inresource-based learning? KOM – Multimedia Communications Lab 21
  22. 22. Objectives1. Investigate ranking algorithms and graph-based recommender techniques for folksonomies2. Design and implement a personalized graph-based recommender system for resource-based learning 1. Identify semantic information to rank learning resources in the application scenario CROKODIL 2. Integrate relevance feedback to personalize ranking of learning resources 3. Integrate explanations for graph-based recommendations KOM – Multimedia Communications Lab 22
  23. 23. Research Approach KOM – Multimedia Communications Lab 23
  24. 24. Current Progress: Conceptual Architecture[Anjorin et al, ISWC Workshop 2011] KOM – Multimedia Communications Lab 24
  25. 25. Current Progress: 3a & 3b 3a 3b 3c 3d KOM – Multimedia Communications Lab 25
  26. 26. Tag Weights based on Semantic Tag Types Tag weights are determined based on the usage frequency of semantic tag types  Tag Types give additional information about the tag and the tag assignment  Assuming usage frequency indicates importance of tag type  Therefore tag types indicate the importance of tags Tag Type Topic Person Goal Event Genre Location Usage Frequency 30% 22% 20% 6% 5% 3% [Böhnstedt, 2011][Anjorin et al, DeLFI Workshop 2011] KOM – Multimedia Communications Lab 26
  27. 27. Graph-Based Recommendations using Semantic Tag Types 0.35 - 0.20 = 0.15 0.32 - 0.20 = 0.11 R1.1 Weight: 0.35 0.20 0.52 - 0.20 = 0.320.06 0.30 0.20 0.05 R1.2.1 Weight: 0.52 0.30 0.22[Anjorin et al, DeLFI Workshop 2011] KOM – Multimedia Communications Lab 27
  28. 28. Graph-Based Recommendations usingSemantic Tag TypesApproach:  Traverse the links between activities to find relevant resources (3-hop transitive associations) [Huang et al, 2004]  Weight resources based on semantic tag types  Rank resources according to resource weights propagated along the activity hierarchyAim:  To generate recommendations for new users (alleviate the cold-start problem)  To alleviate the data sparsity problem[Anjorin et al, DeLFI Workshop 2011] KOM – Multimedia Communications Lab 28
  29. 29. Next StepsAnalyze Ranking Algorithms for FolksonomiesInvestigate Semantic Information in the application scenarioCROKODILImplementation of ConceptsEvaluation of Concepts KOM – Multimedia Communications Lab 29
  30. 30. Overview1. Basics2. Application Scenario: CROKODIL3. Related Work  Ranking Algorithms in Folksonomies  Recommender Systems in E-learning4. Research Topic  Research Question  Objectives  Research Approach  Current Progress5. Future Work6. Summary KOM – Multimedia Communications Lab 30
  31. 31. Future Work: 3c & 3d 3a 3b 3c 3d KOM – Multimedia Communications Lab 31
  32. 32. Future Work: Recommendation Feedback Loop Recommendation Feedback Loop  Rank Resources  Explain recommendations  Relevance Feedback from Learner  Re-Rank Resources 1. Rank 2. Explain Feedback[Harrach and Anjorin, ITiCSE 2011] KOM – Multimedia Communications Lab 32
  33. 33. Future Work: Evaluations 5 & 6 KOM – Multimedia Communications Lab 33
  34. 34. Overview1. Basics2. Application Scenario: CROKODIL3. Related Work  Ranking Algorithms in Folksonomies  Recommender Systems in E-learning4. Research Topic  Research Question  Objectives  Research Approach  Current Progress5. Future Work6. Summary KOM – Multimedia Communications Lab 34
  35. 35. Personalized Recommender Systems forResource-Based LearningMotivation  Due to the vast amount of resources 0.35 - 0.20 = 0.15 0.32 - 0.20 = 0.11 available on the Internet, learners R1.1 Weight: 0.35 0.20 require support in identifying and 0.52 - 0.20 = 0.32 0.06 ranking relevant resources for 0.30 learning purposes. 0.05 0.20 R1.2.1 Weight: 0.52 0.30Challenge 0.22  Exploit additional semantic Contributions information in folksonomies to  Approach using activity hierarchies improve graph-based and semantic tag types to rank recommendations learning resources  Identify semantic information in a  Conceptual architecture of a resource-based learning scenario personalized recommender system like CROKODIL, which could be providing explanations and used to rank learning resources considering relevance feedback from the learner KOM – Multimedia Communications Lab 35
  36. 36. Publications[ARS11] Mojisola Anjorin, Christoph Rensing, Ralf Steinmetz: Towards Ranking in Folksonomies for Personalized Recommender Systems in E-Learning (accepted for publication). October 2011.[ARB+11] Mojisola Anjorin, Christoph Rensing, Kerstin Bischoff, Christian Bogner, Lasse Lehmann, Anna Lenka Reger, Nils Faltin, Achim Steinacker, Andy Lüdemann, Renato Domínguez García: CROKODIL - a Platform for Collaborative Resource-Based Learning (accepted for publication). September 2011.[RBP+11] Christoph Rensing, Christian Bogner, Thomas Prescher, Renato Domínguez García, Mojisola Anjorin: Aufgabenprototypen zur Unterstützung der Selbststeuerung im Ressourcen-basierten Lernen. DeLFI 2011, Sept 2011.[ABR11] Mojisola Anjorin, Doreen Böhnstedt, Christoph Rensing: Towards Graph-Based Recommendations for Resource-Based Learning using Semantic Tag Types. DeLFI 2011, Sept 2011.[AaaC11] Mojisola Anjorin, Renato Domínguez García, Christoph Rensing: CROKODIL: a platform supporting the collaborative management of web resources for learning purposes. ITiCSE, ACM, June 2011.[HA11] Sebastian Harrach, Mojisola Anjorin: Optimizing collaborative learning processes by using recommendation systems. ITiCSE, ACM, June 2011.[Ren11-2] Christoph Rensing,Stephan Tittel, Mojisola Anjorin: Location based Learning Content Authoring and Content Access in the docendo platform. PerCom-WORKSHOPS 2011, March 2011. KOM – Multimedia Communications Lab 36
  37. 37. Literature[ Abel et al, 2008] Fabian Abel, Nicola Henze, and Daniel Krause. Analyzing Ranking Algorithms in Folksonomy Systems. Technical Report, 2008.[Böhnstedt, 2011] Doreen Böhnstedt. Phd Thesis, Technische Universität Darmstadt, 2011.[Huang et al, 2004] Zan Huang, Hsinchun Chen, and Daniel Zeng. Applying Associative Retrieval Techniques to Alleviate the Sparsity Problem in Collaborative Filtering. ACM Transactions of Information Systems, 2004.[Hotho et al, 2006] Andreas Hotho, Robert Jäschke, Christoph Schmitz, and Gerd Stumme. Information Retrieval in Folksonomies: Search and Ranking. In ESWC, Lecture Notes in Computer Science, 2006.[Manouselis et al, 2011] Nikos Manouselis, Hendrik Drachsler, Riina Vuorikari, Hans G. K. Hummel, and Rob Koper. Recommender Systems in Technology Enhanced Learning. In Recommender Systems Handbook. Springer, 2011.[Peters, 2010] Isabella Peters. Folksonomies. Indexing and Retrieval in Web 2.0. De Gruyter - Saur, Berlin, 2010. KOM – Multimedia Communications Lab 37
  38. 38. Any Questions ? KOM – Multimedia Communications Lab 38
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×