# Jonathan Halcrow

Jonathan received his Bachelor's in Math and Physics from Georgia Tech in 2003, then his PhD in Physics from Georgia Tech in 2008. In his PhD work, he studied turbulence in Plane Couette Flow.
At Google, his main interests are in graph algorithms for semi-supervised learning, how to optimize graph construction for particular learning tasks, and graph neural networks.

### Research Areas

Authored Publications

Google Publications

Other Publications

Sort By

Stars: Tera-Scale Graph Building for Clustering and Learning

Warren Schudy

NeurIPS 2022 (2022)

Preview abstract
A fundamental procedure in the analysis of massive datasets is the construction of similarity graphs. Such graphs play a key role for many downstream tasks, including clustering, classification, graph learning, and nearest neighbor search. For these tasks, it is critical to build graphs which are sparse yet still representative of the underlying data. The benefits of sparsity are twofold: firstly, constructing dense graphs is infeasible in practice for large datasets, and secondly, the runtime of downstream tasks is directly controlled by the sparsity of the similarity graph. In this work, we present Stars: a highly scalable method for building extremely sparse graphs via two-hop spanners, which are graphs where similar points are connected by a path of length at most two. Stars can construct two-hop spanners with significantly fewer similarity comparisons, which are a major bottleneck for learning based models where comparisons are expensive to evaluate. Theoretically, we demonstrate that Stars builds a graph in nearly-linear time, where approximate nearest neighbors are contained within two-hop neighborhoods. To complement our results, we have deployed Stars for multiple data sets allowing for graph building at the Tera-Scale, i.e., for graphs with hundreds of billions of nodes and tens of trillions of edges. We evaluate the performance of Stars for clustering and graph learning, and demonstrate 10~1000-fold improvements in pairwise similarity comparisons and significant improvements in runtime with negligible quality loss.
View details

Grale: Designing Networks for Graph Learning

Alexandru Moșoi

Sam Ruth

Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, Association for Computing Machinery (2020), 2523–2532

Preview abstract
How can we find the right graph for semi-supervised learning? In real world applications, the choice of which edges to use for computation is the first step in any graph learning process. Interestingly, there are often many types of similarity available to choose as the edges between nodes, and the choice of edges can drastically affect the performance of downstream semi-supervised learning systems. However, despite the importance of graph design, most of the literature assumes that the graph is static.
In this work, we present Grale, a scalable method we have developed to address the problem of graph design for graphs with billions of nodes. Grale operates by fusing together different measures of (potentially weak) similarity to create a graph which exhibits high task-specific homophily between its nodes. Grale is designed for running on large datasets. We have deployed Grale in more than 20 different industrial settings at Google, including datasets which have tens of billions of nodes, and hundreds of trillions of potential edges to score. By employing locality sensitive hashing techniques, we greatly reduce the number of pairs that need to be scored, allowing us to learn a task specific model and build the associated nearest neighbor graph for such datasets in hours, rather than the days or even weeks that might be required otherwise.
We illustrate this through a case study where we examine the application of Grale to an abuse classification problem on YouTube with hundreds of million of items. In this application, we find that Grale detects a large number of malicious actors on top of hard-coded rules and content classifiers, increasing the total recall by 89% over those approaches alone.
View details

No Results Found