Towards Democratizing AI: Scaling & Learning (Fair) Graph Representations in an Implementation Agnostic Fashion

12-1 pm ET, Tue. 30 January 2024, ITE 325b

Srinivasan Parthasarathy, Ohio State University

Recently there has been a surge of interest in designing graph embedding methods. Few, if any, can scale to a large-sized graph with millions of nodes due to both computational complexity and memory requirements. In this talk, I will present an approach to redress this limitation by introducing the MultI-Level Embedding (MILE) framework -- a generic methodology allowing con-temporary graph embedding methods to scale to large graphs. MILE repeatedly coarsens the graph into smaller ones using a hybrid matching technique to maintain the backbone structure of the graph. It then applies existing embedding methods on the coarsest graph and refines the embeddings to the original graph through a graph convolution neural network that it learns. Time permitting, I will then describe one of several natural extensions to MILE  -  in a distributed setting (DistMILE) to further improve the scalability of graph embedding or mechanisms -  to learn fair graph representations (FairMILE).  

The proposed MILE framework and variants (DistMILE, FairMILE), are agnostic to the underlying graph embedding techniques and can be applied to many existing graph embedding methods without modifying them and is agnostic to their implementation language. Experimental results on five large-scale datasets demonstrate that MILE significantly boosts the speed (order of magnitude) of graph embedding while generating embeddings of better quality, for the task of node classification. MILE can comfortably scale to a graph with 9 million nodes and 40 million edges, on which existing methods run out of memory or take too long to compute on a modern workstation. Our experiments demonstrate that DistMILE learns representations of similar quality with respect to other baselines while reducing the time of learning embeddings even further (up to 40 x speedup over MILE). FairMILE similarly learns fair representations of the data while reducing the time of learning embeddings.

Undergraduate or Graduate Research Assistant 

Description: seeking an undergraduate or graduate research assistant to work on a NASA funded project at the University of Maryland, Baltimore County.  Through this project, you will interact with organizations in the academic, government, and private sectors in South America, Spain, and Portugal, to assess their use of satellite data and other NASA assets. You will assist with 1) research to identify key organizations and experts 2) stakeholder interviews (conducted in Spanish) and survey data collection and analysis and 3) writing articles for magazines and peer-reviewed journals in English and Spanish.

Required

-       Undergraduate or graduate degree (or working towards a degree) in sociology, geography, social data science, public policy or related field.

-       Native Spanish fluency, both oral and written

Preferred

-       Experience with survey and interview data analysis

-       Experience with quantitative and statistical data analysis

-       Lived, studied, or worked in Latin America and/or Spain.  

Time Commitment: 15-20 hours/week

For more information contact Ana Prados at aprados@umbc.edu or 703-338-2119

Please send CVs to aprados@umbc.edu