Big Graph Data Science

One of the challenges in big data analytics lies in being able to reason collectively about extremely large, heterogeneous, incomplete, noisy interlinked data.  We need data science techniques which can
represent and reason effectively with this form of rich and multi-relational graph data. 

In this presentation, Dr. Lise Getoor will describe some common collective inference patterns needed for graph data including: collective classification (predicting missing labels for nodes in a network), link prediction (predicting potential edges), and entity resolution (determining when two nodes refer to the same underlying
entity).  Dr. Getoor will describe two key capabilities required, relational feature construction and collective inference, and briefly describe some of the cutting edge analytic tools being developed within the machine learning, AI, and database communities.

Speaker: Lise Getoor, UC Santa Cruz