1 / 30

Social Media Recommendation based on People and Tags (SIGIR2010)

Social Media Recommendation based on People and Tags (SIGIR2010). IBM Research Lab Ido Guy,Naama Zwerdling Inbal Ronen,David Carmel,Erel Uziel. Social Networks and Discovery( SaND ). Direct entity-entity relations. Item Recommendation Widget. Result. Result. Result.

phong
Download Presentation

Social Media Recommendation based on People and Tags (SIGIR2010)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Social Media Recommendation based on People and Tags(SIGIR2010) IBM Research Lab IdoGuy,NaamaZwerdling InbalRonen,DavidCarmel,ErelUziel

  2. Social Networks and Discovery(SaND)

  3. Direct entity-entity relations

  4. Item Recommendation Widget

  5. Result

  6. Result

  7. Result

  8. On Social Networks and Collaborative Recommendation(SIGIR2009) Dept. of Computing Science University of Glasgow United Kingdom LoannisKonstas, Joemon M Jose Vassilios Stathopoulos

  9. Collaborative Filtering

  10. Random Walk with Restarts How closely related are two nodes in a graph？ a = in every step there is a probability q = is a column vector of zeros with the element corresponding to the starting node set to 1 P(t) = denotes the probability that the random walk at step t

  11. Social Graph and the Sub-matrices

  12. Result

  13. Graph-based Recommendation on Social Networks(IEEE2010) School of Electronics Engineering and Computer Science Peking University Beijing, P.R. China Ziqi Wang, Yuwei Tan, Ming Zhang

  14. Tag-based Promotion Algorithms

  15. User’s similarity based on their tagging information

  16. Result

  17. Result RWR = Random Walk with Restart RWUR = User-Resource Tag-based Algorithm RWUU = User-User Tag-based Algorithm P@k = Precision at rank K S@k = Success at rank K

  18. Learning Optimal Ranking with Tensor Factorization for Tag Recommendation(KDD2009) University of Hildesheim Steffen Rendle, Leandro BalbyMarinho AlexandrosNanopoulos, Lars Schmidt-Thieme

  19. Ternary Relationship

  20. 0/1 Interpretation Scheme The semantics are obviously incorrect Also from a sparsity point of view the 0/1 scheme leads to a problem

  21. Post-based Ranking Interpretation Scheme

  22. Tensor Factorization Model U = Users T = Tags I = Items C = Small core tensor k = are the dimensions of the low-rank approximation

  23. Datasets

  24. Result

  25. Result

  26. Predicting Tie Strength With Social Media(SIGCHI2009) University of Illinois at Urbana-Champaign Eric Gillbert, Karrie Karahalios

  27. Strong Ties and Weak Ties • Strong ties are the people you really trust, people whose social circles tightly overlap with your own. • Weak ties, conversely, are merely acquaintances

  28. Seven Dimensions On FaceBook

  29. Error Analysis Interviews • Asymmetric Friendships • Educational Difference • Confounding the Medium

  30. Limitations • Limitations • We purposely worked from theory to extend this research beyond just Facebook. • The specific predictive variable coefficients may not move beyond Facebook. • Conclusion • We have revealed a specific mechanism by which tie strength manifests itself in social media.

More Related