BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Relational Machine Learning for Knowledge Graphs. - Maximilian Nic
 kel\, MIT
DTSTART:20150310T110000Z
DTEND:20150310T120000Z
UID:TALK58329@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:Knowledge graphs\, which store facts in form of entities and t
 heir relationships\, have found important applications in areas such as We
 b search and question answering. Recently\, machine learning for knowledge
  graphs has received considerable attention as it can be used to discover 
 previously unknown facts\, to categorize and disambiguate entities\, and t
 o support automated knowledge base construction. However\, knowledge graph
 s pose also unique challenges for machine learning\, due to their size and
  their complex\, relational structure.\nIn this talk\, I will present a no
 vel latent factor model for knowledge graphs (and graph-structured data in
  general) that achieves both\, state-of-the-art relational learning and hi
 gh scalability. The proposed approach exploits relational information thro
 ugh its latent variable structure and allows to create statistical models 
 of entire graphs. To estimate its parameters efficiently\, the model can b
 e cast as a tensor factorization problem whose computational complexity sc
 ales linearly with the size of the data. This enables its application to k
 nowledge graphs consisting of millions of entities and billions of known f
 acts. The proposed approach can be applied to a wide range of tasks includ
 ing link prediction\, entity disambiguation\, and link-based clustering\, 
 for which I will demonstrate its state-of-the-art performance on various b
 enchmark datasets. In addition\, I will briefly discuss how the model enab
 les probabilistic queries on knowledge graphs and how it can be combined w
 ith observable variable models to further increase its predictive performa
 nce and scalability.\n
LOCATION:Small Lecture Theatre\, Microsoft Research Ltd\, 21 Station Road\
 , Cambridge\, CB1 2FB
END:VEVENT
END:VCALENDAR