Updated version of the RecSys TEL lecture I already gave as invited talk in the UK, NL and DE. The conclusion parts is totally new and aligned to the new book on RecSys for Learning at Springer that will appear soon in 2012.

1. Recommender Systems
for Learning
12. 04. 2012 Advanced SIKS course on Technology-Enhanced Learning
Landgoed Huize Bergen, Vught, Nederland
Hendrik Drachsler
Centre for Learning Sciences and Technology (CELSTEC)
Open University of the Netherlands 1

2. Goals of the lecture
1. Crash course Recommender Systems (RecSys)
2. Overview of RecSys in TEL
3. Conclusions and open research issues
for RecSys in TEL
2

3. Introduction into
Recommender Systems
Introduction Objectives
Technologies
Evaluation
3

4. Introduction::Application areas
Application areas
• E-commerce websites (Amazon)
• Video, Music websites (Netflix, last.fm)
• Content websites (CNN, Google News)
• Other Information Systems (Zite APP)
Major claims
• Highly application-oriented research area, every domain and
task needs a specific RecSys
• Always build around content or products they never
exist on their own
4

5. Introduction::Definition
Using the opinions of a community of users to
help individuals in that community to identify more
effectively content of interest from a potentially
overwhelming set of choices.
Resnick & Varian (1997). Recommender Systems, Communications of the ACM, 40(3).
5

6. Introduction::Definition
Using the opinions of a community of users to
help individuals in that community to identify more
effectively content of interest from a potentially
overwhelming set of choices.
Resnick & Varian (1997). Recommender Systems, Communications of the ACM, 40(3).
Any system that produces personalized
recommendations as output or has the effect of
guiding the user in a personalized way to interesting
or useful objects in a large space of possible options.
Burke R. (2002). Hybrid Recommender Systems: Survey and Experiments,
User Modeling & User Adapted Interaction, 12, pp. 331-370.
5

7. Introduction::Example
6

8. Introduction::Example
6

9. Introduction::Example
6

10. Introduction::Example
6

11. Introduction::Example
6

12. Introduction::Example
6

13. Introduction::Example
6

14. Introduction::Example
6

15. Introduction::Example
What did we learn from the small exercise?
• There are different kinds of recommendations
a. People who bought X also bought Y
b. There are options to receive even more personalized
recommendations
• When registering, we have to tell the RecSys what we like
(and what not). Thus, it requires information to offer suitable
recommendations and it learns our preferences.
6

16. Introduction:: The Long Tail
Anderson, C. (2004). The Long Tail. Wired Magazine.
7

17. Introduction:: The Long Tail
“We are leaving the age of information and
entering the age of recommendation”.
Anderson, C. (2004)
Anderson, C. (2004). The Long Tail. Wired Magazine.
7

18. Introduction::Emergence
Johnson, S. (2001). Emergence. New York Scribner.
8

19. Introduction::Emergence
Johnson, S. (2001). Emergence. New York Scribner.
8

20. Introduction::Emergence
Johnson, S. (2001). Emergence. New York Scribner.
8

21. Introduction::Emergence
Johnson, S. (2001). Emergence. New York Scribner.
8

22. Introduction::Emergence
Johnson, S. (2001). Emergence. New York Scribner.
8

23. Introduction:: Age of RecSys?
...10 minutes on Google.

24. Introduction:: Age of RecSys?
...10 minutes on Google.

25. Introduction:: Age of RecSys?
... another 10 minutes, research on RecSys is
becoming very popular.
Some examples:
– ACM RecSys conference
– ICWSM: Weblog and Social Media
– WebKDD: Web Knowledge Discovery and Data Mining
– WWW: The original WWW conference
– SIGIR: Information Retrieval
– ACM KDD: Knowledge Discovery and Data Mining
– LAK: Learning Analytics and Knowledge
– Educational data mining conference
– ICML: Machine Learning
– ...
... and various workshops, books, and journals.
10

26. Objectives
of RecSys probabilistic combination of
– Item-based method
– User-based method
– Matrix Factorization
– (May be) content-based method
The idea is to pick from my
previous list 20-50 movies that
share similar audience with
“Taken”, then how much I will like
depend on how much I liked those
early movies
– In short: I tend to watch this movie
because I have watched those
movies … or
11
– People who have watched those
movies also liked this movie

27. Objectives::RecSys Aims
• Converting Browsers into
Buyers
• Increasing Cross-sales
• Building Loyalty
Foto by markhillary
Schafer, Konstan & Riedel, (1999). RecSys in e-commerce. Proc. of the 1st ACM on
electronic commerce, Denver, Colorado, pp. 158-169.
12

28. Objectives::RecSys Tasks
Find good items
presenting a ranked list of
recommendendations.
probabilistic combination of
– Item-based method
– User-based method
– Matrix Factorization
– (May be) content-based method
Find all good items
user wants to identify all
The idea is to pick from my
items that might be previous list 20-50 movies that
share similar audience with
interesting, e.g. medical “Taken”, then how much I will like
depend on how much I liked those
or legal cases early movies
– In short: I tend to watch this movie
because I have watched those
Herlocker, Konstan, Borchers, & Riedl (2004). Evaluating Collaborative Filtering
movies … or
Recommender Systems. ACM Transactions on–Informationhave watched those pp. 5-53.
13
People who Systems, 22(1),
movies also liked this movie

29. Objectives::RecSys Tasks
Find good items Receive sequence of items
presenting a ranked list of sequence of related items is
recommendendations. recommended to the user,
e.g. music recommender
probabilistic combination of
– Item-based method
– User-based method
– Matrix Factorization
Find all good items Annotation in context
– (May be) content-based method
user wants to identify all predicted usefulness of an
items that might be item that pick from mythatis currently
The idea is to the user
previous list 20-50 movies
interesting, e.g. medical viewing, e.g. linkslike
share similar audience with within a
“Taken”, then how much I will
or legal cases websitehow much I liked those
depend on
early movies
– In short: I tend to watch this movie
because I have watched those
Herlocker, Konstan, Borchers, & Riedl (2004). Evaluating Collaborative Filtering
movies … or
Recommender Systems. ACM Transactions on–Informationhave watched those pp. 5-53.
13
People who Systems, 22(1),
movies also liked this movie

30. Objectives::RecSys Tasks
Find good items Receive sequence of items
presenting a ranked list of sequence of related items is
recommendendations. recommended to the user,
e.g. music recommender
There are more tasks available... of
probabilistic combination
– Item-based method
– User-based method
– Matrix Factorization
Find all good items Annotation in context
– (May be) content-based method
user wants to identify all predicted usefulness of an
items that might be item that pick from mythatis currently
The idea is to the user
previous list 20-50 movies
interesting, e.g. medical viewing, e.g. linkslike
share similar audience with within a
“Taken”, then how much I will
or legal cases websitehow much I liked those
depend on
early movies
– In short: I tend to watch this movie
because I have watched those
Herlocker, Konstan, Borchers, & Riedl (2004). Evaluating Collaborative Filtering
movies … or
Recommender Systems. ACM Transactions on–Informationhave watched those pp. 5-53.
13
People who Systems, 22(1),
movies also liked this movie

31. RecSys Technologies
1. Predict how much a user
may like a certain product
2. Create a list of Top-N
best items
3. Adjust its prediction
based on feedback of the
target user and like-
minded users
Hanani et al., (2001). Information Filtering: Overview of Issues, Research and Systems",
User Modeling and User-Adapted Interaction, 11.
14

32. RecSys Technologies
1. Predict how much a user
may like a certain product
2. Create a list of Top-N
best items
3. Adjust its prediction
based on feedback of the Just some examples
target user and like- there are more
minded users technologies available.
Hanani et al., (2001). Information Filtering: Overview of Issues, Research and Systems",
User Modeling and User-Adapted Interaction, 11.
14

33. Technologies::Collaborative filtering
User-based filtering
(Grouplens, 1994)
Take about 20-50 people who share
similar taste with you, afterwards
predict how much you might like an
item depended on how much the others
liked it.
You may like it because your
“friends” liked it.
15

34. Technologies::Collaborative filtering
User-based filtering Item-based filtering
(Grouplens, 1994) (Amazon, 2001)
Take about 20-50 people who share Pick from your previous list 20-50 items
similar taste with you, afterwards that share similar people with “the
predict how much you might like an target item”, how much you will like the
item depended on how much the others target item depends on how much the
liked it. others liked those earlier items.
You may like it because your You tend to like that item because
“friends” liked it. you have liked those items.
15

35. Technologies::Content-based filtering
Information needs of user and characteristics of items are
represented in keywords, attributes, tags that describe
past selections, e.g., TF-IDF.
16

36. Technologies::Hybrid RecSys
Combination of techniques to overcome
disadvantages and advantages of single techniques.
Advantages Disadvantages
probabilistic combination of
– Item-based method
• No content analysis • Cold-start problem
– User-based method
– Matrix Factorization
• Quality improves • Over-fitting
– (May be) content-based method
• No cold-start problem • New user / item problem
The idea is to pick from my
• No new user / item • Sparsity
previous list 20-50 movies that
share similar audience with
problem “Taken”, then how much I will like
depend on how much I liked those
early movies
– In short: I tend to watch this movie
because I have watched those
movies … or
17
– People who have watched those
movies also liked this movie

37. Technologies::Hybrid RecSys
Combination of techniques to overcome
disadvantages and advantages of single techniques.
Advantages Disadvantages
probabilistic combination of
– Item-based method
• No content analysis • Cold-start problem
– User-based method
– Matrix Factorization
• Quality improves • Over-fitting
– (May be) content-based method
• No cold-start problem • New user / item problem
The idea is to pick from my
• No new user / item • Sparsity
previous list 20-50 movies that
share similar audience with
problem “Taken”, then how much I will like
Just some examples there
depend on how much I liked those
early movies
are more (dis)advantages
– In short: I tend to watch this movie
because I have watched those
17
movies … or
available.
– People who have watched those
movies also liked this movie

38. Technologies::Overview
probabilistic combination of
– Item-based method
– User-based method
– Matrix Factorization
– (May be) content-based method
Hanani et al., (2001). Information Filtering: Overview of Issues, Research and Systems",
User Modeling and User-Adapted Interaction, 11, 2001
18

39. Evaluation
of RecSys
probabilistic combination of
– Item-based method
– User-based method
– Matrix Factorization
– (May be) content-based method
The idea is to pick from my
previous list 20-50 movies that
share similar audience with
“Taken”, then how much I will like
depend on how much I liked those
early movies
– In short: I tend to watch this movie
because I have watched those
movies … or
19
– People who have watched those
movies also liked this movie

40. Evaluation::General idea
Most of the time based on performance measures
(“How good are your recommendations?”)
For example:
•Predict what rating will a user give an item?
•Will the user select an item?
•What is the order of usefulness of items to a user?
Herlocker, Konstan, Riedl (2004). Evaluating Collaborative Filtering Recommender
Systems. ACM Transactions on Information Systems, 22(1), 5-53.
20

41. Evaluation::Reference datasets
... and various commercial datasets.
21

42. Evaluation::Approaches
Measures 1. Offline study
•User preference
•Prediction accuracy
•Coverage
•Confidence
•Trust
•Novelty 2. User study
•Serendipity
•Diversity
•Utility
•Risk
•Robustness +
•Privacy
•Adaptivity
•Scalability
22

43. Evaluation::Metrics
Precision – The portion of
recommendations that were
successful. (Selected by the
algorithm and by the user)
Recall – The portion of relevant
items selected by algorithm
compared to a total number of
relevant items available.
F1 - Measure balances Precision
and Recall into a single
measurement.
Gunawardana, A., Shani, G., (2009). A Survey of Accuracy Evaluation Metrics of
Recommendation Tasks, Journal of Machine Learning Research, 10(Dec):2935−2962,
2009. 23

44. Evaluation::Metrics
Precision – The portion of
recommendations that were
successful. (Selected by the
algorithm and by the user)
Recall – The portion of relevant
items selected by algorithm
compared to a total number of
relevant items available.
F1 - Measure balances Precision
and Recall into a single
measurement.
Gunawardana, A., Shani, G., (2009). A Survey of Accuracy Evaluation Metrics of
Recommendation Tasks, Journal of Machine Learning Research, 10(Dec):2935−2962,
2009. 23

45. Evaluation::Metrics
Precision – The portion of
recommendations that were
successful. (Selected by the
algorithm and by the user)
Recall – The portion of relevant
items selected by algorithm
compared to a total number of
relevant items available.
F1 - Measure balances Precision
and Recall into a single
measurement.
Gunawardana, A., Shani, G., (2009). A Survey of Accuracy Evaluation Metrics of
Recommendation Tasks, Journal of Machine Learning Research, 10(Dec):2935−2962,
2009. 23

46. Evaluation::Metrics
Precision – The portion of
recommendations that were
successful. (Selected by the
algorithm and by the user)
Recall – The portion of relevant
items selected by algorithm
compared to a total number of
relevant items available.
F1 - Measure balances Precision Just some examples there
and Recall into a single are more metrics available
measurement. like MAE, RSME.
Gunawardana, A., Shani, G., (2009). A Survey of Accuracy Evaluation Metrics of
Recommendation Tasks, Journal of Machine Learning Research, 10(Dec):2935−2962,
2009. 23

47. Evaluation::Metrics
5
Conclusion:
4
Pearson is better
RMSE
than Cosine, 3
Pearson
because less 2
errors in predicting Cosine
1
TOP-N items. 0
Netflix BookCrossing
Gunawardana, A., Shani, G., (2009). A Survey of Accuracy Evaluation Metrics of
Recommendation Tasks, Journal of Machine Learning Research, 10(Dec):2935−2962,
2009. 24

48. Evaluation::Metrics
5
Conclusion:
4
Pearson is better
RMSE
than Cosine, 3
Pearson
because less 2
errors in predicting Cosine
1
TOP-N items. 0
Netflix BookCrossing
News Story Clicks
Conclusion: 80%
Cosine better than Precision
60%
Pearson, because
40%
of higher precision
20%
and recall value on
TOP-N items. 0%
5% 10% 15% 20% 25% 30% 35% 40%
Recall
Gunawardana, A., Shani, G., (2009). A Survey of Accuracy Evaluation Metrics of
Recommendation Tasks, Journal of Machine Learning Research, 10(Dec):2935−2962,
2009. 24

49. RecSys::TimeToThink
What do you expect that a RecSys for
Learning should do with respect to ...
• Objectives
• Tasks
• Technology
Blackmore’s custom-built LSD Drive
• Evaluation http://www.flickr.com/photos/
rootoftwo/
25

50. Goals of the lecture
1. Crash course Recommender Systems (RecSys)
2. Overview of RecSys in TEL
3. Conclusions and open research issues
for RecSys in TEL
26

51. Recommender Systems
for TEL
Introduction Objectives
Technologies
Evaluation
27

52. TEL RecSys::Definition
Using the experiences of a community of
learners to help individual learners in that
community to identify more effectively learning
content or peer students from a potentially
overwhelming set of choices.
Extended definition by Resnick & Varian (1997). Recommender Systems, Communications
of the ACM, 40(3).
28

53. F
TEL RecSys::Learning spectrum
M
W
Cross, J., Informal learning. Pfeifer. (2006).
29

54. The Long Tail
Graphic: Wilkins, D., (2009). 30

55. The Long Tail of Learning
Graphic: Wilkins, D., (2009). 30

56. The Long Tail of Learning
Formal
Informal
Graphic: Wilkins, D., (2009). 30

57. TEL RecSys::Technologies
31

58. TEL RecSys:: Technologies
Drachsler, H., Pecceu, D., Arts, T., Hutten, E., Rutledge, L., Van Rosmalen, P., Hummel, H. G. K., & Koper, R.
(2009). ReMashed - Recommendations for Mash-Up Personal Learning Environments. In U. Cress, V.
Dimitrova & M. Specht (Eds.), Learning in the Synergy of Multiple Disciplines. Proceedings of the EC-
TEL 2009 (pp. 788-793). September, 29 - October, 2, 2009, Nice, France. Springer LNCS Vol. 5794.
32

59. TEL RecSys:: Technologies
Drachsler, H., Pecceu, D., Arts, T., Hutten, E., Rutledge, L., Van Rosmalen, P., Hummel, H. G. K., & Koper, R.
(2009). ReMashed - Recommendations for Mash-Up Personal Learning Environments. In U. Cress, V.
Dimitrova & M. Specht (Eds.), Learning in the Synergy of Multiple Disciplines. Proceedings of the EC-
TEL 2009 (pp. 788-793). September, 29 - October, 2, 2009, Nice, France. Springer LNCS Vol. 5794.
33

60. TEL RecSys:: Technologies
RecSys Task:
Find good items
Hybrid RecSys:
•Content-based on
interests
•Collaborative filtering
Drachsler, H., Pecceu, D., Arts, T., Hutten, E., Rutledge, L., Van Rosmalen, P., Hummel, H. G. K., & Koper, R.
(2009). ReMashed - Recommendations for Mash-Up Personal Learning Environments. In U. Cress, V.
Dimitrova & M. Specht (Eds.), Learning in the Synergy of Multiple Disciplines. Proceedings of the EC-
TEL 2009 (pp. 788-793). September, 29 - October, 2, 2009, Nice, France. Springer LNCS Vol. 5794.
33

61. TEL RecSys::Tasks
Find good items
e.g. relevant items for a learning
task or a learning goal
The idea is to pick from my
previous list 20-50 movies that
share similar audience with
“Taken”, then how much I will like
depend on how much I liked those
early movies
– In short: I tend to watch this movie
Drachsler, H., Hummel, H., Koper, R., (2009). Identifyinghave goal, user model and
because I the watched those
conditions of recommender systems for formal and or
movies … informal learning. Journal of
Digital Information. 10(2). 34
– People who have watched those
movies also liked this movie

62. TEL RecSys::Tasks
Find good items
e.g. relevant items for a learning
task or a learning goal
Receive sequence of items
e.g. recommend a learning path
to achieve a certain
competence
The idea is to pick from my
previous list 20-50 movies that
share similar audience with
“Taken”, then how much I will like
depend on how much I liked those
early movies
– In short: I tend to watch this movie
Drachsler, H., Hummel, H., Koper, R., (2009). Identifyinghave goal, user model and
because I the watched those
conditions of recommender systems for formal and or
movies … informal learning. Journal of
Digital Information. 10(2). 34
– People who have watched those
movies also liked this movie

63. TEL RecSys::Tasks
Find good items
e.g. relevant items for a learning
task or a learning goal
Receive sequence of items
e.g. recommend a learning path
to achieve a certain
competence
Annotation in context The idea is to pick from my
e.g. take into account location, previous list 20-50 movies that
share similar audience with
time, noise level, prior “Taken”, then how much I will like
knowledge, peers around depend on how much I liked those
early movies
– In short: I tend to watch this movie
Drachsler, H., Hummel, H., Koper, R., (2009). Identifyinghave goal, user model and
because I the watched those
conditions of recommender systems for formal and or
movies … informal learning. Journal of
Digital Information. 10(2). 34
– People who have watched those
movies also liked this movie

64. Evaluation
of TEL
RecSys probabilistic combination of
– Item-based method
– User-based method
– Matrix Factorization
– (May be) content-based method
The idea is to pick from my
previous list 20-50 movies that
share similar audience with
“Taken”, then how much I will like
depend on how much I liked those
early movies
– In short: I tend to watch this movie
because I have watched those
movies … or
35
– People who have watched those
movies also liked this movie

65. TEL RecSys::Review study
36

66. TEL RecSys::Review study
Manouselis, N., Drachsler, H., Vuorikari, R., Hummel, H. G. K., & Koper, R. (2011).
Recommender Systems in Technology Enhanced Learning. In P. B. Kantor, F. Ricci,
L. Rokach, & B. Shapira (Eds.), Recommender Systems Handbook (pp. 387-415).
Berlin: Springer. 36

67. TEL RecSys::Review study
Manouselis, N., Drachsler, H., Vuorikari, R., Hummel, H. G. K., & Koper, R. (2011).
Recommender Systems in Technology Enhanced Learning. In P. B. Kantor, F. Ricci,
L. Rokach, & B. Shapira (Eds.), Recommender Systems Handbook (pp. 387-415).
Berlin: Springer. 36

68. TEL RecSys::Review study
Conclusions:
Half of the systems (11/20) still at design or prototyping
stage only 9 systems evaluated through trials with
human users.
Manouselis, N., Drachsler, H., Vuorikari, R., Hummel, H. G. K., & Koper, R. (2011).
Recommender Systems in Technology Enhanced Learning. In P. B. Kantor, F. Ricci,
L. Rokach, & B. Shapira (Eds.), Recommender Systems Handbook (pp. 387-415).
Berlin: Springer. 36

69. The TEL recommender
research is a bit like this...
37

70. The TEL recommender
research is a bit like this...
We need to design for each domain an
appropriate recommender system that fits the goals, tasks,
and particular constraints
37

71. But...
“The performance results
of different research
efforts in recommender
systems are hardly
comparable.”
(Manouselis et al., 2010)
Kaptain Kobold
http://www.flickr.com/photos/
kaptainkobold/3203311346/
38

72. But...
TEL recommender
experiments lack results
“The performance
transparency and
of different research
efforts in recommender
standardization.
systems are hardly
They need to be
comparable.”
repeatable to test:
•(Manouselis et al., 2010)
Validity
• Verification Kaptain Kobold
http://www.flickr.com/photos/
• Compare results
kaptainkobold/3203311346/
38

73. Data-driven Research and Learning Analytics
EATEL-
Hendrik Drachsler (a), Katrien Verbert (b)
(a) CELSTEC, Open University of the Netherlands
(b) Dept. Computer Science, K.U.Leuven, Belgium
39

75. TEL RecSys::Evaluation/datasets
41

76. TEL RecSys::Evaluation/datasets
Drachsler, H., Bogers, T., Vuorikari, R., Verbert, K., Duval, E., Manouselis, N., Beham, G.,
Lindstaedt, S., Stern, H., Friedrich, M., & Wolpers, M. (2010). Issues and Considerations
regarding Sharable Data Sets for Recommender Systems in Technology Enhanced Learning.
Presentation at the 1st Workshop Recommnder Systems in Technology Enhanced Learning
(RecSysTEL) in conjunction with 5th European Conference on Technology Enhanced
Learning (EC-TEL 2010): Sustaining TEL: From Innovation to Learning and Practice.
September, 28, 2010, Barcelona, Spain. 41

77. 42

78. Evaluation::Metrics
MAE – Mean Absolute Error:
Deviation of recommendations
from the user-specified ratings.
The lower the MAE, the more
accurately the RecSys predicts user
ratings.
Verbert, K., Drachsler, H., Manouselis, N., Wolpers, M., Vuorikari, R., Beham, G., Duval, E.,
(2011). Dataset-driven Research for Improving Recommender Systems for Learning. Learning
Analytics & Knowledge: February 27-March 1,43 2011, Banff, Alberta, Canada

79. Evaluation::Metrics
MAE – Mean Absolute Error:
Deviation of recommendations
from the user-specified ratings.
The lower the MAE, the more
accurately the RecSys predicts user
ratings.
Outcomes:
Tanimoto similarity +
item-based CF was
the most accurate.
Verbert, K., Drachsler, H., Manouselis, N., Wolpers, M., Vuorikari, R., Beham, G., Duval, E.,
(2011). Dataset-driven Research for Improving Recommender Systems for Learning. Learning
Analytics & Knowledge: February 27-March 1,43 2011, Banff, Alberta, Canada

80. Evaluation::Metrics
MAE – Mean Absolute Error:
Deviation of recommendations
from the user-specified ratings.
The lower the MAE, the more
accurately the RecSys predicts user
ratings.
Outcomes:
•User-based CF Algorithm that
predicts the top 10 most relevant
Outcomes:
items for a user has a F1 score
Tanimoto similarity +
of almost 30%.
item-based CF was
•the most accurate.
Implicit ratings like download
rates, bookmarks can
successfully be used in TEL.
Verbert, K., Drachsler, H., Manouselis, N., Wolpers, M., Vuorikari, R., Beham, G., Duval, E.,
(2011). Dataset-driven Research for Improving Recommender Systems for Learning. Learning
Analytics & Knowledge: February 27-March 1,43 2011, Banff, Alberta, Canada

81. Goals of the lecture
1. Crash course Recommender Systems (RecSys)
2. Overview of RecSys in TEL
3. Conclusions and open research issues
for RecSys in TEL
44

82. 10 years of TEL RecSys research in one book
Chapter 1: Background
Chapter 2: TEL context
Recommender
Chapter 3: Extended survey Systems for
of 42 RecSys Learning
Chapter 4: Challenges and
Outlook
Manouselis, N., Drachsler, H., Verbert, K., Duval, E.
(2012). Recommender Systems for Learning. Berlin:
Springer.
45

83. 10 years of TEL RecSys research in one book
Chapter 1: Background
Chapter 2: TEL context
Recommender
Chapter 3: Extended survey Systems for
of 42 RecSys Learning
Chapter 4: Challenges and
Outlook
Manouselis, N., Drachsler, H., Verbert, K., Duval, E.
(2012). Recommender Systems for Learning. Berlin:
Springer.
45

84. A framework for TEL RecSys
46

85. Analysis according to the framework
Supported tasks
47

86. Analysis according to the framework
Supported tasks
Domain model
47

87. Analysis according to the framework
Supported tasks
Domain model
User model
47

88. Analysis according to the framework
Supported tasks
Domain model
User model
Personalization Approach
47

89. Available TEL datasets
48

90. TEL RecSys::Open issues
1. Evaluation
2. Datasets
3. Context
4. Visualization
5. Virtualization
6. Privacy
49

91. TEL RecSys::Ideal research design
1. A selection of datasets
for your RecSys task
2. An offline study of different
algorithms on the datasets
3. A comprehensive controlled user study
to test psychological, pedagogical
and technical aspects
4. Rollout of the RecSys in
real-life scenarios
50

92. Thank you for attending this lecture!
This silde is available at:
http://www.slideshare.com/Drachsler
Email: hendrik.drachsler@ou.nl
Skype: celstec-hendrik.drachsler
Blogging at: http://www.drachsler.de
Twittering at: http://twitter.com/HDrachsler
51

93. TEL RecSys::TimeToThink
• Consider the Recommender System
framework and imagine some great TEL
RecSys that could support you in your
stakeholder role
alternatively
• Name a learning task where a TEL
RecSys would be useful for.
52