BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Weak Signals from the Web: New Perspectives in Collaborative Filte
 ring
DTSTART:20170602T140000
DTEND:20170602T160000
DTSTAMP:20260510T115022Z
UID:d65d5d3f7873b1a00f69060b7ec43242135336315aa5a514c453b47f
CATEGORIES:Conferences - Seminars
DESCRIPTION:Jérémie Rappaz\nEDIC candidacy exam\nExam president: Prof. R
 obert West\nThesis advisor: Prof. Karl Aberer\nCo-examiner: Prof. Pierre D
 illenbourg\n\nAbstract\nPersonalized recommendation has become a crucial a
 spect of many information systems\, as it increases content visibility\, c
 reates users engagement and facilitates navigation. With their growing pre
 sence and the broadening of their applications\, recommender systems are e
 xposed to an ever-growing amount of users. \nTraditional methods\, relyin
 g on user ratings as a primary source of information\, are unable to produ
 ce recommendations for people that have never evaluated any products. \nB
 ased on this observation\, new methods have been proposed to provide accur
 ate results from observations coming from users' natural activity. In this
  work\, we discuss three examples of recommender systems that rely on user
 s implicit feedbacks and take advantage of external sources of information
 . We first discuss an optimization procedure to infer a preference structu
 re from positive-only user interactions. We then discuss a method to rely 
 on the social graph to circumvent high data sparsity. Last\, we examine a 
 method to alleviate the cold-start problem by using visual features.\n\nBa
 ckground papers\nBPR: Bayesian personalized ranking from implicit feedback
 \, Rendle et al.\n[UAI 09] VBPR: Visual Bayesian Personalized Ranking from
  Implicit Feedback\, He et al.\n[AAAI 16] Recommender Systems with Social 
 Regularization\, Ma et al. [WSDM 11]
LOCATION:CO 121 https://plan.epfl.ch/theme/generalite_thm_plan_public?lang
 =en&room=CO%20121&dim_floor=1&dim_lang=en&tree_groups=centres_nevralgiques
 %2Cacces%2Cmobilite_reduite%2Censeignement%2Ccommerces_et_services%2
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR