Optimizing Recommendation in Collaborative E- Learning by Exploring - - PowerPoint PPT Presentation

▶
optimizing recommendation in collaborative e learning by
SMART_READER_LITE
LIVE PREVIEW

Optimizing Recommendation in Collaborative E- Learning by Exploring - - PowerPoint PPT Presentation

Aug 06, 2023 33 likes •307 views

Workshop eliciting Adaptive Sequences for Learning (WeASeL) UQAM, Montral (Canada), 12 June 2018 Optimizing Recommendation in Collaborative E- Learning by Exploring DBpedia and Association Rules 1 Samia Beldjoudi 2 Hassina Seridi 1

slide-1
SLIDE 1

1Samia Beldjoudi
 2Hassina Seridi

Workshop eliciting Adaptive Sequences for Learning (WeASeL) 
 UQAM, Montréal (Canada), 12 June 2018

Optimizing Recommendation in Collaborative E- Learning by Exploring DBpedia and Association Rules

ITS 2018

1Superior School of Industrial Technologies, Annaba, Algeria 1, 2Laboratory of Electronic Document Management LabGED Badji

Mokhtar University, Annaba, Algeria

1s.beldjoudi@epst-annaba.dz 2Seridi@labged.net

slide-2
SLIDE 2

Work plan

General Context

Experiment

Conclusion

Contribution

Motivations

2

slide-3
SLIDE 3

General Context: Social Web

3

Is a set of social relations that link people through the World Wide Web

slide-4
SLIDE 4

General Context: Collaborative E-learning

4

slide-5
SLIDE 5

General Context: Folksonomies

5

Indexing systems produced within internet communities

slide-6
SLIDE 6

General Context: 


Recommendation and Collaborative E-learning

6 suggest items: movies, music

  • r products by analyzing what

the users with similar tastes have chosen in the past

slide-7
SLIDE 7

Issues in Folksonomies

  • Tag ambiguity (Polysemy: many sense):

7

Apple

slide-8
SLIDE 8

Example

8

slide-9
SLIDE 9

Recommender system issue: Diversity and Novelty

  • Accuracy vs Diversity and Novelty in

Recommendation:

9

slide-10
SLIDE 10

Linked Open Data (LOD)

10

Linked data Public & under an open license. Linked Open Data

slide-11
SLIDE 11

Research Question

11

How using LOD to improve recommendation when searching personalized and relevant resources within social E-learning applications?

slide-12
SLIDE 12

Main Contributions

12

Contribution Using LOD to ensure diversity and novelty in recommendation Reduce tag ambiguity problem in recommendation

slide-13
SLIDE 13

Approach description

  • Formally:

a folksonomy is a tuple F = <L, T, R, A> L : learners T : tags R : resources A : the relationships between the three preceding elements, i.e. A ⊆ U x T x R

13

slide-14
SLIDE 14

Approach description

➔ Extacting 3 Social networks: ✓ network relating tags and users, ✓ network relating tags and resources ✓ network relating users and resources. ➔ We represent these social networks by three matrices LT, RT, RL:

  • LT = [Xij] where : Xij
  • RT = [Yij] where: Yij
  • RL = [Zij] where: Zij

14

slide-15
SLIDE 15

Association Rules

15

slide-16
SLIDE 16

Associations Rules and Folksonomies


16

Transaction-id ➔ Learner Transaction items ➔ tags used by the learner

Learner (transaction-id) Tags (itemsets) L1 Software,…………………………., Java L2 Software, ………………………… L3 Java, ……………………….,Software L4 ……………..,Java L5 Java,……………………………,Software Software ⇒ Java

slide-17
SLIDE 17

Steps

  • (1)
  • We

test wheth er the tags which are in the antec edent

  • f the

rule are used by the curre nt learn

  • (2)

resou rces tagge d with each tag found in the conse quent

  • f the

rule are candi date to be reco mme nded by the syste m.

  • (3)

the probl ems

  • f tag

ambig uity: calcul ate simila rities betwe en learn ers.

17

slide-18
SLIDE 18

Example

18

Computer Informatics Mac Java Sun R1 R2 R3

slide-19
SLIDE 19

Example

19

Sun

L1 L2 … L10 L11 L12 Lm R1 R2 R3 T1 T2 T3 … … Tn L1 L2 L10 L11 L12 Lm

R1 R2 R3

L1 L2 … L10 L11 L12 Lm R1 T1 T2 T3 … … Tn L1 L2 L10 L12

slide-20
SLIDE 20

Diversity in Recommendation

20 User profile

Recommendation R e c

  • m

m e n d a t i

  • n
slide-21
SLIDE 21

LOD exploration to insure diversity and novelty

21

Tagged

R4 R1 R3 R2

LOD exploration Recommendation Evaluation 


slide-22
SLIDE 22

1712 tag assignment s 150 users 543 tags 744 resources

Del.icio.us database

22

120 association rules (support= 0.5 and confidence = 0.6. computer ⇒ programming: 60% of the users using the tag ''computer'' also use the tag ''programming''.

slide-23
SLIDE 23

Evaluation Methodology

23

Select a set of tags containing ambiguous tags: (114 tags) We randomly removed sets of resources tagged by these ambiguous tags. Repeat this process five times for each tag in order to make a cross- validation.

slide-24
SLIDE 24

Experimental Results

24

  • Deviation value:

The averages are very promising for the community in general ➔ the small values of standard deviations indicate that the metrics are also promising for each user individually.

Precision Recall F1 Diversity Novelty 0.78 0.71 0.74 0.76 1.2 Precision Recall F1 Diversity Novelty 0.15 0.09 0.1 0.2 0.34

slide-25
SLIDE 25

Conclusion

25

slide-26
SLIDE 26

Future work

  • Ant Colony Optimization

(ACO) Algorithm • Event detection

26

slide-27
SLIDE 27

Thanks…

27

s.beldjoudi@epst-annaba.dz