Heterogeneous Graph Neural Network Research Articles

Towards human-like perception: Learning structural causal model in heterogeneous graph

Heterogeneous graph neural networks have become popular in various domains. However, their generalizability and interpretability are limited due to the discrepancy between their inherent inference flows and human reasoning logic or underlying causal relationships for the learning problem. This study introduces a novel solution, HG-SCM (Heterogeneous Graph as Structural Causal Model). It can mimic the human perception and decision process through two key steps: constructing intelligible variables based on semantics derived from the graph schema and automatically learning task-level causal relationships among these variables by incorporating advanced causal discovery techniques. We compared HG-SCM to seven state-of-the-art baseline models on three real-world datasets, under three distinct and ubiquitous out-of-distribution settings. HG-SCM achieved the highest average performance rank with minimal standard deviation, substantiating its effectiveness and superiority in terms of both predictive power and generalizability. Additionally, the visualization and analysis of the auto-learned causal diagrams for the three tasks aligned well with domain knowledge and human cognition, demonstrating prominent interpretability. HG-SCM’s human-like nature and its enhanced generalizability and interpretability make it a promising solution for special scenarios where transparency and trustworthiness are paramount.

The application of heterogeneous graph neural network and deep reinforcement learning in hybrid flow shop scheduling problem

The hybrid flow shop scheduling problem (HFSP) is a fundamental optimization problem in the process industries. HFSP needs to consider both job selection and machine allocation. The existing HFSP solution methods represented by deep reinforcement learning (DRL) are challenging to extract complex state features and lack certain adaptability in dealing with large-scale HFSP. To solve HFSP with a large scale and relatively strong flexibility, we propose an end-to-end DRL architecture based on the heterogeneous graph neural network to learn an optimal scheduling policy. First of all, by adding a set of machine nodes, the disjunctive graph of the traditional flexible scheduling problem is extended and transformed into a heterogeneous graph, which intuitively represents the allocation relationship between operations and machines, and it is regarded as the scheduling state at different times; secondly, a new HFSP heterogeneous graph neural network (HGNN) algorithm is established to embed potential relationships between different types of nodes and edges, and an embedding of the entire graph is generated; finally, we use the graph-level representation vectors of heterogeneous graphs as input to the proximal policy optimization (PPO) algorithm for learning the optimal scheduling policy. The experimental results show that compared with several well-known heuristic algorithms, the scheduling policy obtained by the architecture through learning can have good generalization performance in simple and complex problems, respectively, and has a relatively high solution efficiency.

An integrated control strategy for simultaneous robot assignment, tool change and preventive maintenance scheduling using Heterogeneous Graph Neural Network

There has been a leap in the field of smart manufacturing with the advancement of automation systems, robotic technology, big data analytics, and state-of-the-art Artificial Intelligence (AI) and Machine Learning (ML) algorithms. Three very important aspects of smart manufacturing systems are system productivity, product quality, and maintenance of machines and equipment. These three issues are strongly interrelated and collectively determine the performance of a smart manufacturing system. Although there has been significant studies in production control, quality control and maintenance scheduling to address each of these aspects individually, there has been a lack of sufficient studies taking all of them into consideration in one control scheme. In this paper, a mobile multi-skilled robot operated Flexible Manufacturing System (FMS) is considered and a model that integrates robots, individual workstation processes and product quality is developed using a Heterogeneous Graph Structure. Heterogeneous Graph Neural Network (HGNN) is used to aggregate local information from different nodes of the graph model to create node embeddings that represent global information. A control problem is then formulated in the Decentralized Partially Observable Markov Decision Process (Dec-POMDP) framework to simultaneously consider robot assignment, product quality and maintenance scheduling. The problem is solved using Multi-Agent Reinforcement Learning (MARL). A case study is presented to demonstrate the effectiveness of the HGNN-MARL control strategy by comparing it to three baselines and the naive MARL policy without HGNN.

MECCH: Metapath Context Convolution-based Heterogeneous Graph Neural Networks

Heterogeneous graph neural networks (HGNNs) were proposed for representation learning on structural data with multiple types of nodes and edges. To deal with the performance degradation issue when HGNNs become deep, researchers combine metapaths into HGNNs to associate nodes closely related in semantics but far apart in the graph. However, existing metapath-based models suffer from either information loss or high computation costs. To address these problems, we present a novel Metapath Context Convolution-based Heterogeneous Graph Neural Network (MECCH). MECCH leverages metapath contexts, a new kind of graph structure that facilitates lossless node information aggregation while avoiding any redundancy. Specifically, MECCH applies three novel components after feature preprocessing to extract comprehensive information from the input graph efficiently: (1) metapath context construction, (2) metapath context encoder, and (3) convolutional metapath fusion. Experiments on five real-world heterogeneous graph datasets for node classification and link prediction show that MECCH achieves superior prediction accuracy compared with state-of-the-art baselines with improved computational efficiency. The code is available at https://github.com/cynricfu/MECCH.

Centrality-based Relation aware Heterogeneous Graph Neural Network

The representation of heterogeneous graph nodes has become a hot research topic due to its diverse applications. However, extant approaches can only give consideration partly to three aspects: node structure, semantics and features. To better integrate these three aspects, a new semi-supervised graph neural network is proposed in this paper, called the CRHGNN (Centrality-based Relation aware Heterogeneous Graph Neural Network). The CRHGNN consists of four components as follows. The first component performs the encoding work and aims to capture the structure and semantics of the nodes. The second and third components perform attention mechanisms and information aggregation, respectively. The CRHGNN learns the mutual attention between nodes and carries out feature learning with less overfitting and fewer oversmoothing problems. The last component performs relation fusion, aiming to obtain a compact representation of the nodes. Experiments are conducted to evaluate the representation learning of nodes on three real-world heterogeneous graph datasets and demonstrate that the proposed model is very competitive in terms of node classification and node clustering tasks.

Heterogeneous graphs neural networks based on neighbor relationship filtering

In recent years, heterogeneous graph neural networks have been applied to the analysis of complex networks, and in ethereum transaction, fraudsters disguise themselves as normal transaction accounts through their behavior, increasing the fault tolerance of using heterogeneous graph neural networks. This paper proposes a heterogeneous graph neural network approach based on neighbor filtering to identify fraudulent ethereum accounts. Specifically, firstly, the collected data of all 386612 ethereum transactions are constructed as transaction subsets G1, G2 and G3. Then, to measure the similarity between ethereum transaction accounts, this paper proposes a similarity measure based on random walks and uses reinforcement learning to find the best neighbor for each relation in the heterogeneous network, and aggregation within and between relations to represent the neighborhood relationship between neighboring nodes. To solve the overfitting problem of inter-relationship aggregation, this paper adds initial residuals to inter-relationship aggregation so that the neighbor aggregation process can be applied to deeper GNNs, resulting in a 2% increase in the AUC metric, which is important for the effective identification of ethereum identities.

Heterogeneous Graph Neural Network With Multi-View Representation Learning

In recent years, graph neural networks (GNNs)-based methods have been widely adopted for heterogeneous graph (HG) embedding, due to their power in effectively encoding rich information from a HG into the low-dimensional node embeddings. However, previous works usually easily fail to fully leverage the inherent heterogeneity and rich semantics contained in the complex local structures of HGs. On the one hand, most of the existing methods either inadequately model the local structure under specific semantics, or neglect the heterogeneity when aggregating information from the local structure. On the other hand, representations from multiple semantics are not comprehensively integrated to obtain node embeddings with versatility. To address the problem, we propose a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Heterogeneous Graph Neural Network</i> for HG embedding <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">within a Multi-View representation learning framework</i> (named MV-HetGNN), which consists of a view-specific ego graph encoder and auto multi-view fusion layer. MV-HetGNN thoroughly learns complex heterogeneity and semantics in the local structure to generate comprehensive and versatile node representations for HGs. Extensive experiments on three real-world HG datasets demonstrate the significant superiority of our proposed MV-HetGNN compared to the state-of-the-art baselines in various downstream tasks, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">e</i> . <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">g</i> . node classification, node clustering, and link prediction.

Higher order heterogeneous graph neural network based on node attribute enhancement

Heterogeneous graph neural networks (HGNNs) have garnered significant attention owing to their ability to capture attribute information from heterogeneous graphs (HGs). However, practical scenarios involving HGs often suffer from missing node attributes. Furthermore, most existing HGNNs have limitations in exploiting node attributes. Specifically, they cannot entirely capture the attributes of higher order neighbors or only use the higher order homogeneous neighbors, thus disregarding the attributes of heterogeneous neighbors. To address these problems, we propose a higher order heterogeneous graph neural network based on heterogeneous node attribute enhancement (HOAE). We first design an attribute-completion strategy using an advanced transformer based self-attention mechanism to fill in the missing attributes. After that, we propose a simple and efficient attribute enhancement strategy based on heterogeneous attributes, empowering HOAE to fully learn the attributes of heterogeneous neighbors. Additionally, meta-path is incorporated to construct a higher order neighbor-based network, enabling effective learning of higher order attributes. Experimental results on three real world datasets demonstrate that HOAE significantly outperforms state-of-the-art methods. The source code of this work is available at https://github.com/FredJDean/HOAE.

Community answer recommendation based on heterogeneous semantic fusion

The community question-answering system has gradually replaced the search engine as the primary way for people to acquire and share knowledge. Users' interactive behavior establishes complex relationships among multiple entities in the question-answering community. With the accumulation of question-and-answer data, retrieving relevant answers to a new question is time-consuming. Some researchers have conducted related research on answer recommendation, mainly dependent on word-based question–answer literal matching, ignoring text semantics and connected entity relationships. In this paper, we propose a Community Answer Recommendation Model Based on Heterogeneous Semantic Fusion, called CARHSF, which combines text semantic and multi-aspect features to enhance answer recommendations. Firstly, we employ the PCFG to build parse trees of text and extract hierarchical phrases through the iterative traversal method. Meanwhile, we utilize phrase embedding to efficiently quantify semantic relatedness. Secondly, we construct a heterogeneous information network to fuse text semantics and entity-relationship features. Finally, we represent entity relations through the heterogeneous graph neural network and supply the recommended answers based on entity correlations. Comparative experiments on three real datasets demonstrate that CARHSF has advantages in answer recommendation tasks.

Dynamic scheduling for flexible job shop with insufficient transportation resources via graph neural network and deep reinforcement learning

The smart workshop is a powerful tool for manufacturing companies to reduce waste and improve production efficiency through real-time data analysis for self-organized production. Automated Guided Vehicles (AGVs) have been widely used for material handling in smart workshop due to their high degree of autonomy, flexibility and powerful end-to-end capability to cope with logistics tasks in production modes such as multiple species and small batch, and mass customization. However, the highly dynamic, complex and uncertain nature of the smart job shop environment makes production scheduling with insufficient transportation resources in mind a challenge. To this end, this paper addresses the dynamic flexible job shop scheduling problem with insufficient transportation resources (DFJSP-ITR), and learn high-quality priority dispatching rule (PDR) end-to-end to minimize makespan by the proposed deep reinforcement learning (DRL) method. To achieve integrated decision making for operation, machine and AGV, an architecture based on heterogeneous graph neural network (GNN) and DRL is proposed. Considering the impact of different AGV distribution methods on the scheduling objective, this paper compares two different AGV distribution methods. Experiments show that the proposed method has superiority and good generalization ability compared with the current PDRs-based methods regardless of the AGV distribution strategy used.

Multi-agent deep reinforcement learning based decision support model for resilient community post-hazard recovery

After a city-scale natural hazard, policymakers should plan sound decisions on the repair sequence to ensure the resilient recovery of the community, which consists of interdependent infrastructures. Stochastic scheduling for repairing interdependent infrastructure systems is a difficult control problem with huge decision spaces. This study proposes a novel decision support model to determine the optimal restoration policies for the purpose of maximizing disaster resilience. A simulation environment is first developed, consisting of hazard intensity assessment, components damage evaluation, system recovery simulation, and resilience quantification. The graph theory is utilized to represent the interdependencies among different systems, and the heterogeneous graph neural network is integrated into this framework to extract the topology and interdependency information of the whole community. The optimal repair policies approximated by neural networks are trained by a multi-agent deep reinforcement learning algorithm, considering uncertainties of the restoration process. The superiority and efficiency of the proposed method are demonstrated through a case study of the Tsinghua University campus, where different decision-making objectives are considered. The results show that the recovery trajectories determined by the proposed model have the highest performance compared to conventional methods. Besides, the proposed methodology based on transfer learning can achieve high computational efficiency for new damage scenarios. This model is promising to be a high-performance, robust decision-support tool for post-hazard repairing decisions.

Encoding feature set information in heterogeneous graph neural networks for game provenance

Encoding feature set information in heterogeneous graph neural networks for game provenance

Summarizing source code through heterogeneous feature fusion and extraction

Code summarization, which seeks to automatically produce a succinct natural-language description to summarize the functionality of source code, plays an essential role in maintaining the software. Currently, plentiful approaches have been proposed to first encode the source code based on its Abstract Syntax Tree (AST), and then decode it into a textual summary. However, most existing works interpret the AST-based syntax structure as a homogeneous graph, without discriminating the different relations between graph nodes (e.g., the parent–child and sibling relations) in a heterogeneous way. To mitigate this issue, this paper proposes HetCoS to extract the syntactic and sequential features of source code by exploring its inherent heterogeneity for code summarization. Specifically, we first build a Heterogeneous Code Graph (HCG) that fuses the syntax structure and code sequence with eight types of edges/relations designed between graph nodes. Moreover, we present a heterogeneous graph neural network for capturing the diverse relations in HCG. The represented HCG is then fed into a Transformer decoder, followed by a multi-head attention-based copying mechanism to support high-quality summary generation. Extensive experiments on the major Java and Python datasets illustrate the superiority of our approach over sixteen state-of-the-art baselines. To promote reproducibility studies, we make the implementation of HetCoS publicly accessible at https://github.com/GJCEXP/HETCOS.

Construction and application of logistics scheduling model based on heterogeneous graph neural network

The core of logistics is scheduling and monitoring. After the modern interprise logistics development concept change, the development prospect of enterprise logistics is more optimistic. Major enterprises have begun to use intelligent logistics scheduling platforms. In order to solve the problem that heterogeneous information fusion is complex in the temporal heterogeneous graphs, this paper proposes to dynamically store and update node representation through an augmented memory matrix in a memory network. At the same time, the model also designs a novel read-write module for the memory matrix, which can effectively capture the timing information in the long interaction sequence and has high flexibility. The model has significantly improved in tasks such as node classification, timing recommendation and visualization. This paper studies the logistics supply chain of modern enterprises and establishes the mathematical model of vehicle scheduling. This paper takes the non-full load scheduling model as the critical research object. Based on the research of logistics supply chain, the vehicle scheduling model is established. The intelligent heuristic algorithm is applied to solve it, and the effective vehicle distribution scheme and driving route are formed. The simulation results show that the approximate Pareto optimal solution obtained by our designed model and algorithm has good robustness. NSGAIIROELSDR can get a better solution in small-scale scheduling. However, in large-scale numerical experiments, the final solution obtained by MOEA/DROELSDR is obviously better than that of NSGAIIROELSDR, and the running time of MOEA/DROELSDR is also shorter. Therefore, we conclude that MOEA/DROELSDR is more suitable for large-scale scheduling, and NSGAIIROELSDR is more suitable for more minor scheduling.

Topic model for long document extractive summarization with sentence-level features and dynamic memory unit

The Transformer-based summarization models rely solely on the attention mechanism for document encoding, making it difficult to accurately capture long-range dependencies in long documents due to the presence of attention redundancy. To address this issue, we propose an extractive summarization framework guided by a topic model (TopicSum) that utilizes a heterogeneous graph neural network to leverage the topic information as document-level features during the sentence selection process, thereby capturing the long-range dependencies among sentences. The sentence-level features in this topic model align with the basic unit of the extractive summarization task. Additionally, a memory mechanism is employed to dynamically store and update the memory module, reducing the potential of repetitive information guiding sentence selection. We evaluated the model on three large document datasets, namely Pubmed, arXiv, and GovReport, and achieved significantly higher Rouge scores than previous works, including extractive and abstractive models. Furthermore, our experiments demonstrate that recent highly regarded large language models such as ChatGPT are insufficient to handle the long document summarization task directly. The proposed approach in this paper exhibits sufficient competitiveness in terms of both generation quality and deployment conditions.

Graph Learning Augmented Heterogeneous Graph Neural Network for Social Recommendation

Social recommendation based on social network has achieved great success in improving the performance of the recommendation system. Since social network (user-user relations) and user-item interactions are both naturally represented as graph-structured data, Graph Neural Networks (GNNs) have thus been widely applied for social recommendation. Despite the superior performance of existing GNNs-based methods, there are still several severe limitations: (i) Few existing GNNs-based methods have considered a single heterogeneous global graph which takes into account user-user relations, user-item interactions, and item-item similarities simultaneously. That may lead to a lack of complex semantic information and rich topological information when encoding users and items based on GNN. (ii) Furthermore, previous methods tend to overlook the reliability of the original user-user relations which may be noisy and incomplete. (iii) More importantly, the item-item connections established by a few existing methods merely using initial rating attributes or extra attributes (such as category) of items, may be inaccurate or sub-optimal with respect to social recommendation. In order to address these issues, we propose an end-to-end heterogeneous global graph learning framework, namely Graph Learning Augmented Heterogeneous Graph Neural Network (GL-HGNN) for social recommendation. GL-HGNN aims to learn a heterogeneous global graph that makes full use of user-user relations, user-item interactions and item-item similarities in a unified perspective. To this end, we design a Graph Learner (GL) method to learn and optimize user-user and item-item connections separately. Moreover, we employ a Heterogeneous Graph Neural Network (HGNN) to capture the high-order complex semantic relations from our learned heterogeneous global graph. To scale up the computation of graph learning, we further present the Anchor-based Graph Learner (AGL) to reduce computational complexity. Extensive experiments on four real-world datasets demonstrate the effectiveness of our model.

RHGNN: Fake reviewer detection based on reinforced heterogeneous graph neural networks

In e-commerce, fake reviewers frequently post fake reviews to mislead consumers into making unwise shopping decisions, seriously affecting customers’ benefits. Graph neural networks (GNNs) have been widely cultivated in fake reviewer detection in recent years, but they exhibit two main limitations. Firstly, most approaches model reviewer detection with a homogeneous graph, ignoring the interdependent clues among reviewers, reviews, and products, making them ineffective against stealthy fake reviewers. Secondly, few works noticed the camouflage behavior of fake reviewers, i.e., fake reviewers disguise themselves as authentic ones by establishing links with genuine reviewers, which extremely weakens the suspiciousness of fake reviewers and undermines detection performance. This paper proposes RHGNN, a robust fake reviewers detection framework based on Reinforced Heterogeneous Graph Neural Networks. Particularly, we first model fake reviewer detection with a heterogeneous graph consisting of reviews, reviewers, and products, to capture stealthy fraud clues by investigating their heterogeneous interactive relations. Subsequently, we present a reinforced neighbor selector (RNS) to sample and aggregate top−k informative neighbors while filtering out the camouflaged relationships. Finally, we propose a self-supervised heterogeneous graph embedding method based on mutual information to learn the discriminative features of fake reviewers. Experimental results show that RHGNN demonstrates promising and robust performance, especially in detecting well-disguised fake reviewers.

Hierarchical Contrastive Learning Enhanced Heterogeneous Graph Neural Network

Heterogeneous graph neural networks (HGNNs) as an emerging technique have shown superior capacity of dealing with heterogeneous information network (HIN). However, most HGNNs follow a semi-supervised learning manner, which notably limits their wide use in reality since labels are usually scarce in real applications. Recently, contrastive learning, a self-supervised method, becomes one of the most exciting learning paradigms and shows great potential when there are no labels. In this paper, we study the problem of self-supervised HGNNs and propose a novel co-contrastive learning mechanism for HGNNs, named HeCo. Different from traditional contrastive learning which only focuses on contrasting positive and negative samples, HeCo employs cross-view contrastive mechanism. Specifically, two views of a HIN (network schema and meta-path views) are proposed to learn node embeddings, so as to capture both of local and high-order structures simultaneously. Then the cross-view contrastive learning, as well as a view mask mechanism, is proposed, which is able to extract the positive and negative embeddings from two views. This enables the two views to collaboratively supervise each other and finally learn high-level node embeddings. Moreover, to further boost the performance of HeCo, two additional methods are designed to generate harder negative samples with high quality. The essence of HeCo is to make positive samples from different views close to each other by cross-view contrast, and learn the factors invariant to two proposed views. However, besides the invariant factors, view-specific factors complementally provide the diverse structure information between different nodes, which also should be contained into the final embeddings. Therefore, we need to further explore each view independently and propose a modified model, called HeCo++. Specifically, HeCo++ conducts hierarchical contrastive learning, including cross-view and intra-view contrasts, which aims to enhance the mining of respective structures. Extensive experiments conducted on a variety of real-world networks show the superior performance of the proposed methods over the state-of-the-arts.

Adversarial Heterogeneous Graph Neural Network for Robust Recommendation

Recommendation systems play a vital role in identifying the hidden interactions between users and items in online social networks. Recently, graph neural networks (GNNs) have exhibited significant performance gains by modeling the information propagation process in graph-structured data for a recommendation. However, existing GNN-based methods do not have broad applicability to heterogeneous graphs that integrate auxiliary data with diverse types. Moreover, graph structures are susceptible to noise and even unnoticed malicious perturbations, as perturbations from connected nodes can create cumulative effects on a target node in the graph. To enhance the robustness and generalization of GNN-based recommendations, we propose a new optimization model named Adversarial Heterogeneous Graph Neural Network for RECommendation (AHGNNRec). First, AHGNNRec learns user and item embeddings by exploring the distinct contributions of various types of interactions between users and items using a hierarchical heterogeneous graph neural network (HGNN). Second, to produce more robust embeddings for recommendations, we employ the adversarial training (AT) method to optimize the HGNN layers. AT is a min-max optimization training process where the generated adversarial fake nodes from normal nodes with intentional perturbations try to maximally deteriorate the recommendation performance. Following this, we learn about these adversarial user or item nodes by minimizing the impact of an additional regularization term for the recommendation. The experimental outcomes on two real-world benchmark datasets demonstrate the effectiveness of AHGNNRec.

SR-HetGNN: session-based recommendation with heterogeneous graph neural network

SR-HetGNN: session-based recommendation with heterogeneous graph neural network