Multimodal personality understanding plays a critical role in human-centered artificial intelligence. Previous work mainly focus on learning rich multimodal representations for video personality understanding. However, they often suffer from potential harm caused by subject bias (e.g., observable age and unobservable mental states), as subjects originate from diverse demographic backgrounds. Learning such spurious associations between multimodal features and traits may lead to unfair personality understanding. In this work, we construct a Structural Causal Model (SCM) to analyze the impact of these biases from a causal perspective, and propose a novel Dual Causal Adjustment Network (DCAN) to mitigate the interference of subject attributes on personality understanding. Specifically, we design a Back-door Adjustment Causal Learning (BACL) module to block spurious correlations from observable demographic factors via a prototype-based confounder dictionary, and subsequently apply a Front-door Adjustment Causal Learning (FACL) module to address latent and unobservable biases through a learned mediator dictionary intervention, thereby achieving causal disentanglement of representations for deconfounded reasoning. Importantly, we construct a Demographic-annotated Multimodal Student Personality (DMSP) dataset to support the analysis and discussion of fairness-related factors. Extensive experiments on the benchmark dataset CFI-V2 and our DMSP dataset demonstrate that DCAN consistently improves prediction accuracy, reaching 92.11% and 92.90%, respectively. Meanwhile, the improvements in the fairness metrics of equal opportunity and demographic parity are 6.57% and 7.97% on CFI-V2, and 15.38% and 20.06% on the DMSP dataset. Our code and DMSP dataset are available at https://github.com/Sabrina-han/DCAN

Accurate trajectory prediction for all agents within complex environments is a crucial step toward realizing autonomous driving navigation. However, this task poses significant challenges due to the uncertainty surrounding the agent's intentions and the intricate road topology. Existing trajectory prediction methods struggle to strike a balance between accuracy and efficiency. To address this challenge, we propose the graph-based trajectory prediction network (DGATP). The model utilizes a two-layer graph representation to capture both the geometric and topological features of the driving environment information and encodes the static and dynamic driving environments hierarchically. An inter-layer network employing an attention mechanism is employed for feature aggregation, leading to improved local-global feature fusion. Furthermore, we introduce a joint prediction framework for all agents in the scenario, which utilizes dynamic weight learning. This adaptive head enhances the model's capacity without increasing its size, thereby maintaining the efficiency of the inference process and leading to accurate and efficient trajectory predictions.

A reliable driving trajectory prediction of surrounding vehicles is an essential reference for decision-making and safe driving of an autonomous vehicle. Although predicting short-term trajectories can be well achieved, it is still very challenging for long-term prediction of trajectories since the prediction space grows exponentially. In this paper, we propose a novel architecture for trajectory prediction from factored intention estimation (I2T), which decouples the trajectory prediction space into a high-level space for intention estimation and a low-level space for motion prediction. The long-term dependencies between intention cues and future motions during driving are naturally extended to the internal sharing mechanism of I2T, leading to improved performance. Furthermore, we design a Prioriformer model to serve as the backbone network for I2T so that it can accurately capture the long-term dependency couplings related to the task of intention estimation or motion prediction. Prioriformer model adopts a personalized normalization method, which facilitates learning latent representations of long-term features and avoids getting stuck on local optimum. A designed multi-scale fusion encoder extracts features from various receptive fields and then learns richer information from the representation subspaces. An efficient non-autoregressive decoder reduces the pressure in long-term prediction of trajectories while avoiding cumulative errors. Experiments on three real-world motion datasets show that I2T can significantly outperform the state-of-the-art.

Network alignment aims at detecting the corresponding entities across multiple networks, which is an essential basis for the fusion and analysis of multiple network information. Moreover, embedding-based network alignment has gradually become one of the promising methods. However, existing methods ignore the confusing selection problem caused by the similarity-orientated principle of network embedding and over-dependence on the hypothesis of structural consistency. In this paper, we propose an end-to-end Graph Alignment Neural Network (GANN) model with graph-to-sequence learning. GANN mainly consists of two modules: Graph encoder and Sequence decoder. In graph encoder module, we present a restricted network embedding method, which can not only capture the local structure and attribute information of nodes but also realize the constraint of node embedding and space reconciliation. In sequence decoder module, we propose a graph-to-sequence learning model to address large graphs' structural consistency hypothesis problem. In this model, an attention-based LSTM mechanism is introduced to infer a node in the source network corresponding to the candidate node sequence in target networks. In this candidate sequence, the correct aligned node is placed at the top. We demonstrate that GANN outperforms the state-of-the-art methods in network alignment tasks on various real-world datasets.

In the task of traffic flow forecasting, various external factors need to be considered to interfere with the flow, such as weather conditions, traffic accidents, emergency events, and Points of Interest (POIs). While capturing the spatio-temporal dependencies, it is essential to effectively capture the external factors. However, existing studies cannot effectively cascade the information contained in these external factors to traffic features, and lack the co-capture of spatio-temporal features. To address these challenges, we present a Knowledge Representation learning-actuated Spatio-Temporal Graph Neural Network (KR-STGNN) for traffic flow prediction. The Gated Feature Fusion Module (GFFM) is utilized to combine the knowledge embedding with the traffic features, and the traffic features are updated adaptively and dynamically according to the importance of external factors. To conduct the co-capture of spatio-temporal dependencies, we subsequently propose a spatio-temporal feature synchronous capture module combining dilation causal convolution with GRU. Experimental results on a real-world traffic dataset demonstrate that KR-STGNN has superior forecasting performances with different prediction steps, especially for short-term prediction.

Accurately forecasting the future motions of road participants is essential for proactive hazard avoidance and safety planning of autonomous vehicles. Existing methods for motion prediction based on probabilistic generative models are limited to low-accuracy likelihood calculations and relatively finite mode distributions. Recent studies show that score-based models can naturally overcome these limitations. In this work, we present a novel paradigm of conditional score-based models for vehicle motion prediction, called Motion-CSM. First, we model scene contextual representations of interaction regions at the feature level via graph convolutional networks. We then interpolate these representations as conditions into the solution process of the continuous-time reverse stochastic differential equation (SDE) to guide trajectory generation, which progressively converts the known prior distributions into multimodal trajectories including the ground truth modes. The designed stacked Transformer structure with dual control conditions is adopted to learn the score function approximation of the Gaussian perturbation kernel. Finally, we develop multiple consistency constraints to align the inference results of Motion-CSM in reverse SDE solving to improve the self-consistency and stability of multimodal trajectory generation. Experimental results on the real-world motion dataset demonstrate that the multimodal forecasting accuracy of Motion-CSM outperforms state-of-the-art methods.

Multiplex networks contain multiple types of relations between nodes, where each relation type is modeled as one layer. In the real world, a relation type may only depend on certain attribute features of nodes. Most existing multiplex network embedding methods only focus on preserving consistent information or complementary information from multiplex networks. However, these methods ignore the dependency between node attributes and the topology of each relation. To address the problem, we propose a model called DAME (Disentangled-based Adversarial Network for Multiplex Network Embedding). We utilize generative adversarial learning to preserve the consistent and complementary information between different relation types. Meanwhile, we develop a disentangled graph convolution network (DGCN) based on disentangled learning, enabling DAME to capture the dependency between node attributes and each relation type. We conduct extensive experiments on five realworld datasets. Experimental results indicate the effectiveness of DAME on link prediction and node classification tasks.

Network Embedding (NE) has emerged as a powerful tool in many applications. Many real-world networks have multiple types of relations between the same entities, which are appropriate to be modeled as multiplex networks. However, at random walk-based embedding study for multiplex networks, very little attention has been paid to the problems of sampling bias and imbalanced relation types. In this paper, we propose an Adaptive Node Embedding Framework (ANEF) based on cross-layer sampling strategies of nodes for multiplex networks. ANEF is the first framework to focus on the bias issue of sampling strategies. Through metropolis hastings random walk (MHRW) and forest fire sampling (FFS), ANEF is less likely to be trapped in local structure with high degree nodes. We utilize a fixed-length queue to record previously visited layers, which can balance the edge distribution over different layers in sampled node sequence processes. In addition, to adaptively sample the cross-layer context of nodes, we also propose a node metric called Neighbors Partition Coefficient (NPC). Experiments on real-world networks in diverse fields show that our framework outperforms the state-of-the-art methods in application tasks such as cross-domain link prediction and mutual community detection.

The relationship extraction and fusion of networks are the hotspots of current research in social network mining. Most previous work is based on single-source data. However, the relationships portrayed by single-source data are not sufficient to characterize the relationships of the real world. To solve this problem, a Semi-supervised Fusion framework for Multiple Network (SFMN), using gradient boosting decision tree algorithm (GBDT) to fuse the information of multi-source networks into a single network, is proposed in this paper. Our framework aims to take advantage of multi-source networks fusion to enhance the accuracy of the network construction. The experiment shows that our method optimizes the structural and community accuracy of social networks which makes our framework outperforms several state-of-the-art methods.

Graph has been widely used for modeling complex relationship datasets in different application fields. Social networks based recommendation system have obtained satisfactory results in Business Intelligence(BI). However, current personalized recommendation methods based on graph structure generally lack interactivity and seldom consider efficient data management. To address these problems, Graph OnLine Analytical Mining (GraphOLAM) is a promising method, which combines OLAP technology with social networks. We first propose an efficient recommendation framework based on GraphOLAM data cube technology for the recommendation in the insurance service. Based on this framework, a new algorithm framework named RU-GOLAM for insurance is proposed, which combines GraphOLAM dimensional aggregation operation and specific recommendation methods. A series of graphs can be generated by GraphOLAM dimensional aggregation operations, which reflect the relationships of nodes under the constraints of different hierarchical dimensions. Node similarities are calculated to generate the Top-N sequential recommendation based on all of these graphs, which can achieve the balance between the topology of the original graph and high-dimensional information of the nodes. Experiments show that our approach outperforms other baseline algorithms on an insurance service dataset.

Network embedding aims to learn a latent representation of each node which preserves the structure information. Many real-world networks have multiple dimensions of nodes and multiple types of relations. Therefore, it is more appropriate to represent such kind of networks as multiplex networks. A multiplex network is formed by a set of nodes connected in different layers by links indicating interactions of different types. However, existing random walk based multiplex networks embedding algorithms have problems with sampling bias and imbalanced relation types, thus leading the poor performance in the downstream tasks. In this paper, we propose a node embedding method based on adaptive cross-layer forest fire sampling (FFS) for multiplex networks (FFME). We first focus on the sampling strategies of FFS to address the bias issue of random walk. We utilize a fixed-length queue to record previously visited layers, which can balance the edge distribution over different layers in sampled node sequences. In addition, to adaptively sample node's context, we also propose a metric for node called Neighbors Partition Coefficient (N P C ). The generation process of node sequence is supervised by NPC for adaptive cross-layer sampling. Experiments on real-world networks in diverse fields show that our method outperforms the state-of-the-art methods in application tasks such as cross-domain link prediction and shared community structure detection.