Linkless Link Prediction via Relational Distillation

Graph Neural Networks (GNNs) have been widely used on graph data and have shown exceptional performance in the task of link prediction. Despite their effectiveness, GNNs often suffer from high latency due to non-trivial neighborhood data dependency in practical deployments. To address this issue, researchers have proposed methods based on knowledge distillation (KD) to transfer the knowledge from teacher GNNs to student MLPs, which are known to be efficient even with industrial scale data, and have shown promising results on node classification. Nonetheless, using KD to accelerate link prediction is still unexplored. In this work, we start with exploring two direct analogs of traditional KD for link prediction, i.e., predicted logit-based matching and node representation-based matching. Upon observing direct KD analogs do not perform well for link prediction, we propose a relational KD framework, Linkless Link Prediction (LLP). Unlike simple KD methods that match independent link logits or node representations, LLP distills relational knowledge that is centered around each (anchor) node to the student MLP. Specifically, we propose two matching strategies that complement each other: rank-based matching and distribution-based matching. Extensive experiments demonstrate that LLP boosts the link prediction performance of MLPs with significant margins, and even outperforms the teacher GNNs on 6 out of 9 benchmarks. LLP also achieves a 776.37x speedup in link prediction inference compared to GNNs on the large scale OGB-Citation2 dataset.

Empowering Graph Representation Learning with Test-Time Graph Transformation

As powerful tools for representation learning on graphs, graph neural networks (GNNs) have facilitated various applications from drug discovery to recommender systems. Nevertheless, the effectiveness of GNNs is immensely challenged by issues related to data quality, such as distribution shift, abnormal features and adversarial attacks. Recent efforts have been made on tackling these issues from a modeling perspective which requires additional cost of changing model architectures or re-training model parameters. In this work, we provide a data-centric view to tackle these issues and propose a graph transformation framework named GTrans which adapts and refines graph data at test time to achieve better performance. We provide theoretical analysis on the design of the framework and discuss why adapting graph data works better than adapting the model. Extensive experiments have demonstrated the effectiveness of GTrans on three distinct scenarios for eight benchmark datasets where suboptimal data is presented. Remarkably, GTrans performs the best in most cases with improvements up to 2.8%, 8.2% and 3.8% over the best baselines on three experimental settings.

Multi-task Self-supervised Graph Neural Networks Enable Stronger Task Generalization

Self-supervised learning (SSL) for graph neural networks (GNNs) has attracted increasing attention from the graph machine learning community in recent years, owing to its capability to learn performant node embeddings without costly label information. One weakness of conventional SSL frameworks for GNNs is that they learn through a single philosophy, such as mutual information maximization or generative reconstruction. When applied to various downstream tasks, these frameworks rarely perform equally well for every task, because one philosophy may not span the extensive knowledge required for all tasks. In light of this, we introduce ParetoGNN, a multi-task SSL framework for node representation learning over graphs. Specifically, ParetoGNN is self-supervised by manifold pretext tasks observing multiple philosophies. To reconcile different philosophies, we explore a multiple-gradient descent algorithm, such that ParetoGNN actively learns from every pretext task while minimizing potential conflicts. We conduct comprehensive experiments over four downstream tasks (i.e., node classification, node clustering, link prediction, and partition prediction), and our proposal achieves the best overall performance across tasks on 11 widely adopted benchmark datasets. Besides, we observe that learning from multiple philosophies enhances not only the task generalization but also the single task performance, demonstrating that ParetoGNN achieves better task generalization via the disjoint yet complementary knowledge learned from different philosophies.

MLPInit: Embarrassingly Simple GNN Training Acceleration with MLP Initialization

Training graph neural networks (GNNs) on large graphs is complex and extremely time consuming. This is attributed to overheads caused by sparse matrix multiplication, which are sidestepped when training multi-layer perceptrons (MLPs) with only node features. MLPs, by ignoring graph context, are simple and faster for graph data, however they usually sacrifice prediction accuracy, limiting their applications for graph data. We observe that for most message passing-based GNNs, we can trivially derive an analog MLP (we call this a PeerMLP) whose weights can be made identical, making us curious about how do GNNs using weights from a fully trained PeerMLP perform? Surprisingly, we find that GNNs initialized with such weights significantly outperform their PeerMLPs for graph data, motivating us to use PeerMLP training as a precursor, initialization step to GNN training. To this end, we propose an embarrassingly simple, yet hugely effective initialization method for GNN training acceleration, called MLPInit. Our extensive experiments on multiple large-scale graph datasets with diverse GNN architectures validate that MLPInit can accelerate the training of GNNs (up to 33X speedup on OGB-products) and often improve prediction performance (e.g., up to 7.97% improvement for GraphSAGE across 7 datasets for node classification, and up to 17.81% improvement across 4 datasets for link prediction on metric Hits@10). Most importantly, MLPInit is extremely simple to implement and can be flexibly used as a plug-and-play initialization method for message passing-based GNNs.

Flashlight: Scalable Link Prediction with Effective Decoders

Link prediction (LP) has been recognized as an important task in graph learning with its broad practical applications. A typical application of LP is to retrieve the top scoring neighbors for a given source node, such as the friend recommendation. These services desire the high inference scalability to find the top scoring neighbors from many candidate nodes at low latencies. There are two popular decoders that the recent LP models mainly use to compute the edge scores from node embeddings: the HadamardMLP and Dot Product decoders. After theoretical and empirical analysis, we find that the HadamardMLP decoders are generally more effective for LP. However, HadamardMLP lacks the scalability for retrieving top scoring neighbors on large graphs, since to the best of our knowledge, there does not exist an algorithm to retrieve the top scoring neighbors for HadamardMLP decoders in sublinear complexity. To make HadamardMLP scalable, we propose the Flashlight algorithm to accelerate the top scoring neighbor retrievals for HadamardMLP: a sublinear algorithm that progressively applies approximate maximum inner product search (MIPS) techniques with adaptively adjusted query embeddings. Empirical results show that Flashlight improves the inference speed of LP by more than 100 times on the large OGBL-CITATION2 dataset without sacrificing effectiveness. Our work paves the way for large-scale LP applications with the effective HadamardMLP decoders by greatly accelerating their inference.

Are Graph Neural Networks Really Helpful for Knowledge Graph Completion?

Knowledge graphs (KGs) facilitate a wide variety of applications due to their ability to store relational knowledge applicable to many areas. Despite great efforts invested in creation and maintenance, even the largest KGs are far from complete. Hence, KG completion (KGC) has become one of the most crucial tasks for KG research. Recently, considerable literature in this space has centered around the use of Graph Neural Networks (GNNs) to learn powerful embeddings which leverage topological structures in the KGs. Specifically, dedicated efforts have been made to extend GNNs, which are commonly designed for simple homogeneous and uni-relational graphs, to the KG context which has diverse and multi-relational connections between entities, by designing more complex aggregation schemes over neighboring nodes (crucial to GNN performance) to appropriately leverage multi-relational information. The success of these methods is naturally attributed to the use of GNNs over simpler multi-layer perceptron (MLP) models, owing to their additional aggregation functionality. In this work, we find that surprisingly, simple MLP models are able to achieve comparable performance to GNNs, suggesting that aggregation may not be as crucial as previously believed. With further exploration, we show careful scoring function and loss function design has a much stronger influence on KGC model performance, and aggregation is not practically required. This suggests a conflation of scoring function design, loss function design, and aggregation in prior work, with promising insights regarding the scalability of state-of-the-art KGC methods today, as well as careful attention to more suitable aggregation designs for KGC tasks tomorrow.

Explaining Graph-level Predictions with Communication Structure-Aware Cooperative Games

Explaining predictions made by machine learning models is important and have attracted an increased interest. The Shapley value from cooperative game theory has been proposed as a prime approach to compute feature importances towards predictions, especially for images, text, tabular data, and recently graph neural networks (GNNs) on graphs. In this work, we revisit the appropriateness of the Shapley value for graph explanation, where the task is to identify the most important subgraph and constituent nodes for graph-level predictions. We purport that the Shapley value is a no-ideal choice for graph data because it is by definition not structure-aware. We propose a Graph Structure-aware eXplanation (GStarX) method to leverage the critical graph structure information to improve the explanation. Specifically, we propose a scoring function based on a new structure-aware value from the cooperative game theory called the HN value. When used to score node importance, the HN value utilizes graph structures to attribute cooperation surplus between neighbor nodes, resembling message passing in GNNs, so that node importance scores reflect not only the node feature importance, but also the structural roles. We demonstrate that GstarX produces qualitatively more intuitive explanations, and quantitatively improves over strong baselines on chemical graph property prediction and text graph sentiment classification.

Imbalanced Graph Classification via Graph-of-Graph Neural Networks

Graph Neural Networks (GNNs) have achieved unprecedented success in learning graph representations to identify categorical labels of graphs. However, most existing graph classification problems with GNNs follow a balanced data splitting protocol, which is misaligned with many real-world scenarios in which some classes have much fewer labels than others. Directly training GNNs under this imbalanced situation may lead to uninformative representations of graphs in minority classes, and compromise the overall performance of downstream classification, which signifies the importance of developing effective GNNs for handling imbalanced graph classification. Existing methods are either tailored for non-graph structured data or designed specifically for imbalance node classification while few focus on imbalance graph classification. To this end, we introduce a novel framework, Graph-of-Graph Neural Networks (G$^2$GNN), which alleviates the graph imbalance issue by deriving extra supervision globally from neighboring graphs and locally from graphs themselves. Globally, we construct a graph of graphs (GoG) based on kernel similarity and perform GoG propagation to aggregate neighboring graph representations, which are initially obtained by node-level propagation with pooling via a GNN encoder. Locally, we employ topological augmentation via masking nodes or dropping edges to improve the model generalizability in discerning topology of unseen testing graphs. Extensive graph classification experiments conducted on seven benchmark datasets demonstrate our proposed G$^2$GNN outperforms numerous baselines by roughly 5\% in both F1-macro and F1-micro scores. The implementation of G$^2$GNN is available at \href{https://github.com/YuWVandy/G2GNN}{https://github.com/YuWVandy/G2GNN}.

Graph-less Neural Networks: Teaching Old MLPs New Tricks via Distillation

Graph Neural Networks (GNNs) have recently become popular for graph machine learning and have shown great results on wide node classification tasks. Yet, GNNs are less popular for practical deployments in the industry owing to their scalability challenges incurred by data dependency. Namely, GNN inference depends on neighbor nodes multiple hops away from the target, and fetching these nodes burdens latency-constrained applications. Existing inference acceleration methods like pruning and quantization can speed up GNNs to some extent by reducing Multiplication-and-ACcumulation (MAC) operations. However, their improvements are limited given the data dependency is not resolved. Conversely, multi-layer perceptrons (MLPs) have no dependency on graph data and infer much faster than GNNs, even though they are less accurate than GNNs for node classification in general. Motivated by these complementary strengths and weaknesses, we bring GNNs and MLPs together via knowledge distillation (KD). Our work shows that the performance of MLPs can be improved by large margins with GNN KD. We call the distilled MLPs Graph-less Neural Networks (GLNNs) as they have no inference graph dependency. We show that GLNN with competitive performance infer faster than GNNs by 146X-273X and faster than other acceleration methods by 14X-27X. Meanwhile, under a production setting involving both transductive and inductive predictions across 7 datasets, GLNN accuracies improve over stand alone MLPs by 12.36% on average and match GNNs on 6/7 datasets. A comprehensive analysis of GLNN shows when and why GLNN can achieve competitive results to GNNs and suggests GLNN as a handy choice for latency-constrained applications.