TITLE:
Scalable Collective Inference in Heterogenous Networks using PSL

ABSTRACT:
One of the challenges in big data analytics lies in being able to
reason collectively about extremely large, incomplete, noisy
interlinked data.  Some of the common collective inference patterns
include: collective classification (predicting missing labels in the
network), link prediction (predicting missing relationships),
community detection (discovering clusters of interlinked entities),
and entity resolution (determining when two references refer to the
same entity).  In this talk I will overview our recent work on
probabilistic soft logic (PSL), a framework for collective,
probabilistic reasoning in relational domains.  PSL is able to reason
holistically about both entity attributes and relationships among the
entities.  The underlying mathematical framework, which we refer to as
a hinge-loss Markov random field, supports extremely efficient
inference.  I will survey several applications of PSL to problems
in computational social science and knowledge graph identification.
Our recent results show that by building on state-of-the-art optimization
methods in a distributed implementation, we can solve large-scale
problems with millions of random variables orders of magnitude more
quickly than existing approaches.

Joint Work with Stephen Bach, Bert Huang, Matthias Broecheler,
Jay Pujara, Hui Miao, Angelika Kimmig, Ben London, Alex Memory, Stanley Kok,
and Shobeir Fahkraei.