DATA-DRIVEN STUDY: AUGMENTING PREDICTION ACCURACY OF RECOMMENDATIONS IN SOCIAL LEARNING PLATFORMS

286 views

Published on

Poster presented at the 25th Benelux Conference on Artificial Intelligence (BNAIC 2013), TUDelft, the Netherlands

0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
286
On SlideShare
0
From Embeds
0
Number of Embeds
36
Actions
Shares
0
Downloads
3
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide

DATA-DRIVEN STUDY: AUGMENTING PREDICTION ACCURACY OF RECOMMENDATIONS IN SOCIAL LEARNING PLATFORMS

  1. 1. 250" 200" 150" MovieLens" degree% OpenScout" 100" MACE" Travel"well" 50" 0" u1" u2" u3" u4" u5" u6" u7" u8" u9" u10" Top)10%central%users% Degree  centrality  of  first  10   central  users   EU  FP7  Open  Discovery   Space  (ODS)   DATA-­‐DRIVEN  STUDY:  AUGMENTING   PREDICTION  ACCURACY  OF   RECOMMENDATIONS  IN  SOCIAL   LEARNING  PLATFORMS   MACE% 0.1" Tanimoto4Jaccard"(CF1)" Loglikelihood"(CF2)" 0.08" Tanimoto3Jaccard"(CF1)" 0.06" Loglikelihood"(CF2)" Euclidean"(CF3)" 0.04" Graph4based"(CF4)" 0.02" Euclidean"(CF3)" Graph3based"(CF4)" 0" 5" 7" 3" 10" 5" 0.2" 0.06" Tanimoto3Jaccard"(CF1)" 0.04" Loglikelihood"(CF2)" Euclidean"(CF3)" Graph3based"(CF4)" 3" 5" 7" 10" size%of%neighborhood%(n)% •  Tanimoto0Jaccard"(CF1)" 0.15" Loglikelihood"(CF2)" 0.1" Euclidean"(CF3)" 0.05" Graph0based"(CF4)" 0" 0" Classical  collaboraHve   filtering  based  on  nearest   neighbors,  beside  a  graph-­‐ based  approach   Three  datasets  similar  to  the   ODS  data  plus  MovieLens   F1@10% F1@10% 0.25" 0.02" •  10" MovieLens% Travel%well% 0.1" 0.08" Offline  data  study   7" size%of%neighborhood%(n)% size%of%neighborhood%(n)% 1-­‐  Conceptual  model   2-­‐  Offline  data  study  ✓   3-­‐  User  study   4-­‐  Go  online   OPEN  UNIVERSITEIT  NEDERLAND   FIRSTNAME.LASTNAME@OU.NL   0.12" 3" Research  methodology   SOUDE  FAZELI   HENDRIK  DRACHSLER   PETER  SLOEP   OpenScout% 0.14" F1@10% F1@10% RQ1:  How  to  generate  more   accurate  recommendaHons  by   taking  into  account  user   interacHons  in  social  learning   pla_orms?   RQ2:  Can  the  use  of  the  graph   walking  algorithms  improve   the  process  of  finding  like-­‐ minded  users  within  a  social   learning  pla_orm?   0.1" 0.09" 0.08" 0.07" 0.06" 0.05" 0.04" 0.03" 0.02" 0.01" 0" 3" 5" 7" 10" size%of%neighborhood%(n)% F1  score  of  different  user-­‐ based  collaboraHve  filtering   algorithms    

×