Abstract:Message Passing Graph Neural Networks (MPGNNs) have emerged as the preferred method for modeling complex interactions across diverse graph entities. While the theory of such models is well understood, their aggregation module has not received sufficient attention. Sum-based aggregators have solid theoretical foundations regarding their separation capabilities. However, practitioners often prefer using more complex aggregations and mixtures of diverse aggregations. In this work, we unveil a possible explanation for this gap. We claim that sum-based aggregators fail to "mix" features belonging to distinct neighbors, preventing them from succeeding at downstream tasks. To this end, we introduce Sequential Signal Mixing Aggregation (SSMA), a novel plug-and-play aggregation for MPGNNs. SSMA treats the neighbor features as 2D discrete signals and sequentially convolves them, inherently enhancing the ability to mix features attributed to distinct neighbors. By performing extensive experiments, we show that when combining SSMA with well-established MPGNN architectures, we achieve substantial performance gains across various benchmarks, achieving new state-of-the-art results in many settings. We published our code at \url{<a class="link-external link-https" href="https://almogdavid.github.io/SSMA/" rel="external noopener nofollow">this https URL</a>}

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is the limitations of the aggregation module in existing Message Passing Graph Neural Networks (MPGNNs), especially the poor performance of sum - based aggregators in handling neighbor feature mixing. Specifically: 1. **Limited ability to mix neighbor features**: Existing sum - based aggregators cannot effectively "mix" features from different neighbors, which limits their performance in downstream tasks. The paper verifies this by defining the "neighbor mixing" property and conducting experiments. 2. **Gap between theory and practice**: Although sum - based aggregators have good separation capabilities in theory, in practical applications, complex aggregation methods usually perform better. The author believes that this gap may be due to the lack of effective neighbor feature mixing ability of sum - based aggregators. To solve these problems, the author introduces a new aggregation method - **Sequential Signal Mixing Aggregation (SSMA)**. SSMA regards neighbor features as two - dimensional discrete signals and enhances the feature - mixing ability through sequential convolution. This method not only improves the performance of the model but also reaches a new state - of - the - art level in multiple benchmark tests. ### Main contributions 1. **Define the neighbor - mixing property**: The author defines the "neighbor - mixing" property and verifies the limitations of sum - based aggregators in this property through experiments. 2. **Propose SSMA**: Introduce a new aggregation module SSMA with a dimension of \( m = O(n^2 d) \), where \( n \) is the number of neighbors and \( d \) is the feature dimension. SSMA effectively enhances the feature - mixing ability by regarding neighbor features as discrete signals and performing sequential convolution. 3. **Optimize and expand SSMA**: Discuss how to stabilize the optimization process of SSMA and show how to extend it to larger graph structures. 4. **Experimental verification**: Through extensive experiments, it is shown that integrating SSMA into existing MPGNN architectures can significantly improve their performance and achieve state - of - the - art results in multiple benchmark tests. In conclusion, this paper aims to improve the representation - learning ability of MPGNNs on graph - structured data by improving the aggregation module, especially in tasks that require effective mixing of neighbor features.

Sequential Signal Mixing Aggregation for Message Passing Graph Neural Networks

Improving Message-Passing GNNs by Asynchronous Aggregation

Message gain aggregation architecture: a scalable graph neural network for combining large-scale neighborhoods

Multi-Track Message Passing: Tackling Oversmoothing and Oversquashing in Graph Learning Via Preventing Heterophily Mixing

Generalization Bounds for Message Passing Networks on Mixture of Graphons

Deep Neural Message Passing with Hierarchical Layer Aggregation and Neighbor Normalization

SIGMA: Similarity-based Efficient Global Aggregation for Heterophilous Graph Neural Networks

Meta-Aggregator: Learning to Aggregate for 1-bit Graph Neural Networks

Convergence of Message Passing Graph Neural Networks with Generic Aggregation On Large Random Graphs

Memory-based Message Passing: Decoupling the Message for Propogation from Discrimination

SlotGAT: Slot-based Message Passing for Heterogeneous Graph Neural Network

Memory-based message passing: decoupling the message for propagation from discrimination

Alleviating Over-Smoothing via Aggregation over Compact Manifolds

IntraMix: Intra-Class Mixup Generation for Accurate Labels and Neighbors

Multi-graph aggregated graph neural network for heterogeneous graph representation learning

Multi-View Metrics Enhanced Heterogeneous Graph Neural Network for Extractive Summarization

MG-GCN: Fast and Effective Learning with Mix-grained Aggregators for Training Large Graph Convolutional Networks

From Stars to Subgraphs: Uplifting Any GNN with Local Structure Awareness

Diversifying Message Aggregation in Multi-Agent Communication via Normalized Tensor Nuclear Norm Regularization

LSPI: Heterogeneous Graph Neural Network Classification Aggregation Algorithm Based on Size Neighbor Path Identification

Generalizing Aggregation Functions in GNNs: Building High Capacity and Robust GNNs Via Nonlinear Aggregation.