Sequential recommendation has emerged as a fundamental task in various domains, aiming to predict a user's next interaction based on historical behavior. Recent advances in deep sequence models, particularly Transformer-based architectures and the more recent Mamba, have substantially pushed the boundaries of sequential modeling performance. However, existing methods still face two critical challenges. First, many current approaches overlook the hierarchical structures and high-order dependencies among items, typically restricting representation learning to conventional Euclidean spaces, which limits their capacity to capture complex relational information. Second, although Mamba excels at long-range dependency modeling, its reliance on static Feed-Forward Networks (FFNs) hinders its ability to dynamically adapt to evolving user preferences across diverse contexts. To address these limitations, we propose a Hyperbolic-Enhanced Mixture-of-Experts Mamba recommender (HM2Rec) for sequential recommendation. HM2Rec first encodes user-item relationships through hyperbolic graph convolution to exploit hierarchical structure more effectively. Then, a Variational Graph Auto-Encoder (VGAE) is employed to reconstruct node embeddings, improving structural robustness. To further enhance sequential modeling, we integrate Rotary Positional Encoding (RoPE) into Mamba to better capture relative position dependencies, and replace the FFN with Mixture-of-Expert (MOE) module, enabling dynamic and personalized expert selection for each token. Our extensive experiments on four widely-used public datasets demonstrate that HM2Rec outperforms several advanced baseline models.

Multi-object tracking (MOT) aims to track multiple objects while maintaining consistent identities across frames of a given video. In unmanned aerial vehicle (UAV) recorded videos, frequent viewpoint changes and complex UAV-ground relative motion dynamics pose significant challenges, which often lead to unstable affinity measurement and ambiguous association. Existing methods typically model motion and appearance cues separately, overlooking their spatio-temporal interplay and resulting in suboptimal tracking performance. In this work, we propose AMOT, which jointly exploits appearance and motion cues through two key components: an Appearance-Motion Consistency (AMC) matrix and a Motion-aware Track Continuation (MTC) module. Specifically, the AMC matrix computes bi-directional spatial consistency under the guidance of appearance features, enabling more reliable and context-aware identity association. The MTC module complements AMC by reactivating unmatched tracks through appearance-guided predictions that align with Kalman-based predictions, thereby reducing broken trajectories caused by missed detections. Extensive experiments on three UAV benchmarks, including VisDrone2019, UAVDT, and VT-MOT-UAV, demonstrate that our AMOT outperforms current state-of-the-art methods and generalizes well in a plug-and-play and training-free manner.

With rapid urbanization in the modern era, traffic signals from various sensors have been playing a significant role in monitoring the states of cities, which provides a strong foundation in ensuring safe travel, reducing traffic congestion and optimizing urban mobility. Most existing methods for traffic time series modeling often rely on the original data modality, i.e., numerical direct readings from the sensors in cities. However, this unimodal approach overlooks the semantic information existing in multimodal heterogeneous urban data in different perspectives, which hinders a comprehensive understanding of traffic signals and limits the accurate prediction of complex traffic dynamics. To address this problem, we propose a novel Multimodal framework, MTP, for urban Traffic Profiling, which learns multimodal features through numeric, visual, and textual perspectives in the frequency domain. The three branches drive a multimodal perspective of traffic signal learning for augmentation, while the frequency learning strategies delicately refine the information for extraction. Specifically, we first conduct the visual augmentation for the traffic time series, which transforms the original modality into periodicity images and frequency images for visual learning. Also, we augment descriptive texts for the traffic time series based on the specific topic, background information and item description for textual learning. To complement the numeric information, we utilize frequency multilayer perceptrons for learning on the original modality. We design a hierarchical contrastive learning on the three branches to fuse the three modalities. Finally, extensive experiments on six real-world datasets demonstrate superior performance compared with the state-of-the-art approaches.

Contrastive learning has emerged as a promising paradigm by inherently generating self-supervised signals and uncovering latent patterns from interaction data to enhance recommendation performance. However, most current graph contrastive learning-based recommendation methods rely on random augmentation strategies,which may disrupt graph structural information and compromise model robustness. In addition, long-tail items suffer from insufficient exposure, making it difficult to learn high-quality feature rep- resentations, ultimately degrading recommendation effectiveness.To overcome these limitations, this paper presents DDGCL, a dual diffusion-based graph contrastive learning method. A contrastive view optimization module is designed, which employs singular value decomposition to perform low-rank approximation on the interaction graph, efficiently extracting global structural features while accelerating the diffusion process. The diffusion model then performs noise addition and denoising on this basis to generate contrastive views that preserve graph structural information. In addition, a method for embedding augmentation designed for long-tail items is proposed. This module utilizes a conditional diffusion model, where global graph information serves as conditional con- straints to guide the denoising process of long-tail items, thereby improving their representation learning. A comprehensive evaluation on multiple public benchmark datasets demonstrates that DDGCL significantly outperforms various baseline models, validating the effectiveness of the proposed approach.

Hypergraph-based distillation methods have been proposed to mitigate the high computational cost of Hypergraph Neural Networks (HGNNs) in modeling high-order relationships. However, most existing methods use static and uniform distillation strategies for all nodes and hyperedges, ignoring their individual characteristics. In addition, they neglect the student model's capability to independently extract useful internal features. As a result, they are not effective in transferring higher-order structural knowledge from the teacher. To overcome these limitations, we propose ARCHER, an Adaptive and Reinforcement-Guided Contrastive HypER graph Distillation framework that enables a lightweight MLP student model to outperform its HGNN teacher model. First, we design an adaptive strategy that leverages node- and hyperedge-level confidence to mediate error guidance from the teacher model. Second, we introduce a contrastive learning module that guides the student to learn from both the teacher's outputs and its own internal representations, producing more expressive embeddings. Finally, we propose a multi-armed bandit-based reinforcement learning module that dynamically balances multiple loss objectives during training. Experiments on six benchmark datasets demonstrate that our method outperforms existing hypergraph distillation methods.

With the rapid growth of multi-modal content on the Web, robust vision-language models are essential for semantic understanding and classification of web images under diverse and dynamic contexts, supporting Web applications such as multimedia search and recommendation. Prompt learning has proven effective for enhancing vision-language models in semantic image classification tasks. However, previous methods often suffer from poor generalization: the learned prompts tend to overfit the base classes seen during training, leading to poor performance on unseen classes and under distribution shifts. This issue is especially challenging in Web-scale data, where new classes emerge and distributions shift dynamically. To address these limitations, we propose PLIKD, a novel prompt learning method that integrates instance-aware knowledge distillation for robust Web-scale semantic image classification. Specifically, PLIKD introduces an instance-aware knowledge extraction module, which leverages multi-modal large language models through a step-by-step strategy to extract external knowledge for each image instance. To incorporate this extracted knowledge, PLIKD further introduces an instance-aware knowledge distillation module, which consists of two key steps: (1) a dual-teacher strategy for robust and informative knowledge distillation, and (2) fine-grained cross-modal alignment via Smooth and Sparse Optimal Transport. Extensive experiments demonstrate that PLIKD significantly improves generalization to both seen and unseen classes, and remains robust under distribution shifts, outperforming existing state-of-the-art methods on Web-scale semantic image classification.

Multi-modal recommendation has become essential with the rapid expansion of online platforms such as e-commerce and video-sharing applications. In this work, we focus on the Multi-Modal New Item Recommendation (MMNIR) problem, where items with multi-modal content but newly introduced items lack interaction history. The MMNIR problem is particularly challenging in two aspects: (1) a large number of new items are created rapidly over time without any interaction data, (2) not all existing interactions are equally useful, and it is non-trivial to identify informative behaviors from users with similar preferences. However, previous methods fail to identify users with similar preferences and to capture informative behaviors from historical data. Furthermore, conventional models primarily rely on simple co-occurring signals, leading to spurious neighbors and neglecting the informative behaviors of truly similar users with consistent preferences. To fill this gap, we propose Joint Similar User Exploration and Informative Behavior Guidance (SuperG) for solving the MMNIR problem. SuperG first proposes a similar user exploration module to identify users with similar preferences to the target user. Then it incorporates an informative behavior mining module to retrieve informative behaviors from both the target user and similar users' histories to support new item recommendation. Finally, SuperG proposes a behavior-guided hybrid recommendation module to incorporate the retrieved behavioral signals to guide the recommendation of new items. Our empirical study on three real datasets demonstrates that SuperG outperforms the state-of-the-art models under the MMNIR setting.

Recently, hypergraph knowledge distillation has been proposed to alleviate the high computational cost of Hypergraph Neural Networks (HGNNs) when modeling high-order relationships in Web-related graph tasks. Its effectiveness primarily depends on the quality of knowledge transferred from the teacher and the representation capability of the student. However, existing methods remain limited on both sides. On the teacher side, most methods typically rely on a single HGNN teacher, which provides limited structural and semantic knowledge, thereby constraining the upper bound of the student's performance. The potential of exploiting multiple teachers in HGNNs remains largely underexplored. On the student side, existing methods ignore the student's capability to capture high-order semantic and structural information beyond simply imitating teacher outputs, leading to limited representation learning. To address these limitations, we propose MARCH, a framework for Multi-TeAcheR Contrastive Hypergraph Distillation, which advances semantic modeling and distillation for Web-scale structured data. Specifically, MARCH proposes a multi-teacher distillation strategy that adaptively transfers complementary knowledge from multiple teachers at both node and hyperedge levels, empowering the student model to learn richer and more discriminative representations and even outperform its teachers. Extensive experiments on six benchmark datasets demonstrate the superior performance of MARCH.

Graph-level anomaly detection (GLAD) is a critical task to identify graphs with abnormal properties in various domains, ranging from fraudulent social networks to malicious botnets on online platforms. The dominant paradigm for existing GLAD detectors has been partially unsupervised, relying on training data composed exclusively of normal samples. However, this partially unsupervised paradigm inevitably requires a costly expert filtering process to ensure the training data is free of anomalies. This creates a significant gap between current approaches and the real-world necessity of a fully unsupervised paradigm, which involves training a model directly on real-world data ''as-is'', with its inherent mix of normal and anomalous samples. To bridge this gap, we incorporate uncertainty learning into GLAD to promote fully unsupervised learning. We propose two frameworks: Score Uncertainty Learning (SUL) and Graph-data Uncertainty Learning (GUL). Specifically, SUL enhances existing GLAD detectors by modeling uncertainty through Gaussian distributions over the detectors' predictions, adaptively attenuating the influence of potential anomalies. GUL is an end-to-end framework that iteratively optimizes anomaly detection and uncertainty modeling via an Expectation-Maximization algorithm. In addition, we develop a dedicated loss that utilizes potential anomalies to enhance the effectiveness and robustness of GUL. Empirical results on sixteen benchmark datasets, covering real-world graphs from social networks and online platforms, demonstrate the superiority of our methods and highlight the promise of incorporating uncertainty into fully unsupervised GLAD.

Web-scale platforms and online services rely on log-based anomaly detection to safeguard availability, latency SLOs, and user experience. In real-world web interactions, system logs often exhibit irregular temporal intervals, bursty densities, and heterogeneous semantics, which pose significant challenges for log anomaly detection. Existing methods such as LSTM and Transformer assume a fixed input window, which conflicts with the inherently irregular nature of system logs. Moreover, most prior works build a single-view representation, overlooking the multi-relational nature of logs. To overcome these challenges, we propose DyLogNet, a dynamic multi-relational graph framework for log anomaly detection. Specifically, this framework constructs a density-aware dynamic graph with variable-length windows, and represents logs from three relational perspectives: temporal co-occurrence, semantic similarity, and anomaly tendency. Next, we design a cross-layer attention mechanism that integrates heterogeneous structures to highlight the most relevant relations and enhance event representations. Furthermore, a cross-snapshot memory injection module updates global memory through a recurrent unit and injects it into current graph representations via an affine transformation, enabling temporal continuity. Experiments on three public log datasets demonstrate that DyLogNet outperforms state-of-the-art methods, especially in few-shot scenarios.

Object recognition is essential for autonomous cars, and the amalgamation of camera and light detection and ranging (LiDAR) sensor data has emerged as a pivotal method for accurate three-dimensional (3D) object recognition. Contemporary algorithms face challenges with fragmented data, high processing costs, insufficient resolution, and limited dynamic information. This study presents a novel approach utilising 3D U-Net deep learning for precise 3D object detection and localisation by integrating camera and LiDAR data. The process involves obtaining and preprocessing camera and LiDAR data, utilising a geometric 3D frustum method to extract 3D information from LiDAR based on 2D camera bounding boxes, and training a You Only Look Once version 4 (YOLO v4) network to recognise these boundaries in camera images. The detected images are combined with LiDAR data, and a deep U-Net network is utilised to define 3D bounding boxes. Performance is assessed at various noise levels (0 %, 1 %, 2 %, 5 %, and 10 %) in the composite images. This method leverages the benefits of both sensors to better object recognition across diverse shapes and sizes, even in challenging situations, signifying a significant progression towards safer and more reliable autonomous vehicles with improved situational awareness in intricate urban environments.

Despite the significant progress of fully-supervised video captioning, zero-shot methods remain much less explored. In this paper, we propose a novel zero-shot video captioning framework named R etrieval- E nhanced T est- T ime A daptation (RETTA), which takes advantage of existing pre-trained large-scale vision and language models to directly generate captions with test-time adaptation. Specifically, we bridge video and text using four key models: a general video-text retrieval model XCLIP, a general image-text matching model CLIP, a text alignment model AnglE, and a text generation model GPT-2, due to their source-code availability. The main challenge is how to enable the text generation model to be sufficiently aware of the content in a given video so as to generate corresponding captions. To address this problem, we propose using learnable tokens as a communication medium among these four frozen models GPT-2, XCLIP, CLIP, and AnglE. Different from the conventional way that trains these tokens with training data, we propose to learn these tokens with soft targets of the inference data under several carefully crafted loss functions, which enable the tokens to absorb video information catered for GPT-2. This adaptation requires only a few iterations ( e.g. , 16) and does not require ground truth data. Extensive experimental on MSR-VTT, MSVD, and VATEX, show absolute 5.1 % ∼ 32.4 % improvements in CIDEr scores compared to several state-of-the-art zero-shot video captioning methods.

Brain graphs (brain connectivity networks) play an important role in modeling the complex structure of the human brain. Furthermore, brain graph learning based on graph neural networks (GNNs) has recently attracted growing interest. Although existing methods have made great progress in brain disorder prediction and pathogenic analysis, there are two key problems: (1) They rarely utilize the pathogenic reason of brain disorders, that is, salient brain regions always lead to abnormal connections between brain regions, to extract critical brain graph information for disorder analysis; (2) Since most of the available brain graph data is limited, how can we improve the performance of brain graph learning models on insufficient training data? Thus, in this paper, we learn brain graph representations for disorder prediction and analyze disorder-specific brain regions and connections from the subgraph perspective. Besides, we introduce the cross-domain brain graph learning framework to alleviate the problem of poor model performance on limited data. To consider the pathogenic reason by brain subgraphs, we first propose the node entropy of brain graphs based on brain graph properties to extract important nodes. We then introduce subgraph information bottleneck to refine the critical subgraph from the rough subgraph generated by these important nodes, recognizing important connections related to disorders. To achieve a better model performance on limited data, we design a cross-domain brain graph learning framework to improve the subgraph extraction model by the meta-learning method. The subgraph extraction model is pre-trained on a large source training dataset and then quickly adapted to target task dataset. Besides, a simple yet effective feature alignment module is applied to mitigate the negative transfer problem for cross-domain datasets. Extensive experimental results, including disorder prediction and pathogenic analysis on real-world neuroimaging data, demonstrate the effectiveness of our method.

Understanding the complex relationships and behavioral preferences among users in social networks is crucial for elucidating the mechanisms of information diffusion. While recent information diffusion prediction methods, empowered by graph neural networks, have advanced the learning of user representations, they rarely exploit spread interaction feedback. This feedback reflects users' interests in engaging with information and serves as a key driver of information diffusion. Moreover, the underutilization of unlabeled data leads to an overreliance on labeled data, consequently constraining the model's self-learning and generalization capabilities. To address these limitations, we propose a novel microscopic diffusion prediction model based on interaction-enhanced graph neural sequence contrastive learning (IEGSCL). Specifically, we construct a triple graph to explore the diversity of user relationships and preferences through the lenses of trust and interaction. A self-supervised graph contrastive learning module is designed to transfer user intents, maximizing the utility of unlabeled data and tackling the feature extraction challenge. Furthermore, we devise an information-driven gating strategy that adaptively modulates the contributions of social and interactive intents to cascade participation, thereby effectively integrating interaction feedback into the cascade modeling. Finally, we employ maximum mean discrepancy (MMD) to enforce distributional consistency between global relationship representations and local cascade encodings. Extensive experiments on four public datasets validate the superior performance of the proposed model over existing baselines.

Traditional methods for ensuring security and privacy face challenges in safeguarding multimedia data within the IoT-edge continuum, as their significant computational demands render them unsuitable for IoT devices with limited resources. Next, we find that the federated learning techique can naturally adapt to edge frameworks and provide effective data security and privacy protection. In this article, we propose FLiForest, an innovative anomaly detection approach that integrates federated learning with the isolation forest algorithm, tailored for the IoT-edge continuum. Specifically, our method designs a three-stage process, including data collection and sampling, model training, and data testing, to joint-train an isolation forest among clients and edge servers. FLiForest facilitates decentralized model training across IoT devices, enhancing data privacy and reducing computational burden, without necessitating the exchange of multimedia data. Through extensive experiments on a variety of multimedia datasets, the efficacy of our method is benchmarked against the state-of-the-art anomaly detection methods, showcasing its superior detection accuracy and robustness in ensuring data privacy and security.

Federated learning (FL) enables edge devices to collaboratively train a global model without sharing raw data. In the edge environment, resource constraints hinder efficient training and aggregation of FL. Although prior studies have established model pruning as a practical strategy to reduce resource demand, the parameter symmetry problem (i.e., permutation-equivalent parameter orderings in neural networks) remains underexplored. Without addressing this symmetry, pruning on heterogeneous clients results in aggregation mismatches and degraded accuracy. In this paper, we propose Attention-guided One-shot Pruning for Federated Learning (FedAOP) to address these challenges. First, we design an attention module that integrates spatial and channel attention to highlight critical spatial responses and evaluate channel importance. Then, leveraging these importance scores, we propose an attentive pruning algorithm to generate client-specific models, thereby reducing resource consumption. Furthermore, we introduce an aggregation algorithm with attention matching, thereby mitigating the adverse effects of parameter symmetry under heterogeneous pruning. We implementFedAOPon a benchmark FL platform. The experimental results on benchmark datasets demonstrate thatFedAOPconsistently outperforms state-of-the-art baselines by up to 11.3% in accuracy while reducing the average model footprint by 32%.

Graph Contrastive Learning (GCL), as a primary paradigm of graph self-supervised learning, spurs a fruitful line of research in tackling the data sparsity issue by maximizing the consistency of user/item embeddings between different augmented views with random perturbations. However, diversity, as a crucial metric for recommendation performance and user satisfaction, has received rather little attention. In fact, there exists a challenging dilemma in balancing accuracy and diversity. To address these issues, we propose a new Graph Contrastive Learning (DivGCL) model for diversifying recommendations. Inspired by the excellence of the determinant point process (DPP), DivGCL adopts a DPP likelihood-based loss function to achieve an ideal trade-off between diversity and accuracy, optimizing it jointly with the advanced Gaussian noise-augmented GCL objective. Extensive experiments on four popular datasets demonstrate that DivGCL surpasses existing approaches in balancing accuracy and diversity, with an improvement of 23.47% at T@20 (abbreviation for trade-off metric) on ML-1M.

Incomplete multi-view multi-label classification aims to accurately predict labels for each sample in the face of some missing views. Due to its widespread presence in real-world scenarios, it has become an extensively researched topic. In addition to the challenges brought by missing views, it also encounters issues caused by redundant views, whose inclusion fails to make a positive contribution to performance. In this paper, we make the first attempt to take advantage of diffusion models to address the missing view problem and design a strategy to identify and remove redundant views. Specifically, we train a diffusion model conditioned on the pseudo-labels to recover information of missing views. The learned diffusion model can carry data distribution knowledge in training split to the data. Regarding redundant identification strategy, it is designed by considering both the additional information of views and the classification difficulty level of samples, thereby adaptively identifying and removing redundant views. We conduct extensive experiments on five datasets, and the proposed method achieves favorable performance against several state-of-the-art methods on the multi-view multi-label classification task.

Sequential recommendation aims to capture the temporal dependencies of items in a user's historical interactions and make recommendations based on this. Previous generative methods addressed the issue of data not directly reflecting user preference uncertainty by modeling the distribution of latent item representations. Diffusion model (DM)-based methods have achieved significant success due to their high-quality generation and stable training. However, they lack satisfactory user sequence representations to guide the generation process, impacting recommendation performance. Moreover, these methods overlook the drawback of slow inference speed, severely limiting their practical value. To obtain effective generative guidance signals and accelerate the recommendation process, we propose DAE4Rec. In this approach, a Graph Auto-Encoder (GAE) is used to obtain interpretable item node representations, revealing global transitions of items that previous methods struggled to uncover. Then, we use it to construct a generative guidance signal with lower coupling and variance for the diffusion model. Additionally, by employing a non-Markov chain derived from the forward diffusion process, it is the first to implement a 'skip-step' reverse process in diffusion model-based methods. And a creatively designed compensator is used to bridge the performance gap caused by 'skip-step'. Extensive experiments on three real-world datasets demonstrate that DAE4Rec outperforms other state-of-the-art generative sequential recommenders.

Alignment techniques have become central to ensuring that Large Language Models (LLMs) generate outputs consistent with human values.However, existing alignment paradigms often model an averaged or monolithic preference, failing to account for the diversity of perspectives across cultures, demographics, and communities.This limitation is particularly critical in health-related scenarios, where plurality is essential due to the influence of culture, religion, personal values, and conflicting opinions.Despite progress in pluralistic alignment, no prior work has focused on health, likely due to the unavailability of publicly available datasets.To address this gap, we introduce VITAL, a new benchmark dataset comprising 13.1K value-laden situations and 5.4K multiplechoice questions focused on health, designed to assess and benchmark pluralistic alignment methodologies.Through extensive evaluation of eight LLMs of varying sizes, we demonstrate that existing pluralistic alignment techniques fall short in effectively accommodating diverse healthcare beliefs, underscoring the need for tailored AI alignment in specific domains.This work highlights the limitations of current approaches and lays the groundwork for developing health-specific alignment solutions.1

Dynamic pricing on two-sided platforms such as Airbnb presents complex challenges due to the heterogeneity of listings, user behaviours, and contextual variables. In this work, we propose a robust and interpretable pricing framework that leverages Large Language Models (LLMs) and prompt engineering to automate the generation of high-level meta-features from unstructured and structured listing data. These meta-features are designed to capture nuanced semantic features that are often overlooked by traditional feature engineering pipelines. We further integrate these representations into a Transformer-based Graph Neural Network (GNN), which models the relational and spatial dependencies between listings in a data-driven and several relation-construction manner. By combining prompt-driven embeddings with graph-aware contextual learning, our framework significantly enhances price recommendation accuracy while offering transparency through assortativity analysis. Extensive experiments on real-world Airbnb datasets demonstrate our approach's performance in both prediction and unseen data across neighbourhoods and output interpretability. This work highlights the potential of unifying LLMs, structured graph learning, and interpretable AI for next-generation dynamic pricing systems.

Spatio-temporal graphs encode dynamic interactions across space and time, but their size and complexity pose challenges for analysis and computation. Graph sparsification provides an effective solution to these issues by reducing the number of edges while preserving the essential structural and dynamic properties of the network. This reduction is crucial for enhancing the interpretability of complex graphs, revealing hidden patterns, and enabling more efficient computational analysis. However, real-world graphs often exhibit continuous spatial and temporal evolution, which most existing sparsification algorithms, primarily designed for static graphs, fail to address. We introduce STGS (Spatio-Temporal Graph Sparsification), a reinforcement learning-based framework for sparsifying spatio-temporal graphs. By learning to prune edges while preserving key spatio-temporal patterns, STGS enables efficient analysis of evolving systems. Experiments on real-world datasets demonstrate that STGS outperforms existing methods in both structural preservation and downstream forecasting tasks.

Financial trading analytics increasingly demands modular, explainable, and adaptive intelligence systems capable of handling volatile market conditions and multimodal data streams. Recent advances in generative Artificial Intelligence (AI), Large Language Models (LLMs), and graph-based representations have enabled the creation of intelligent agents that can reason over complex, interconnected financial data. We introduce GWise, a graph-structured, generative AI-enabled multi-agent framework for real-time financial trading analytics delivered via secure web services. GWise models financial decision making as a directed computational graph of specialized analytical agents, including technical, fundamental, sentiment, and risk analysis crews, whose outputs are orchestrated through a memory-augmented LLM. This graph-structured design enables transparent, adaptive, and explainable trade recommendations that evolve over time. We demonstrate how the framework's agent orchestration forms a dynamic service graph, facilitating composability, fault isolation, and scalable deployment through cloud-native APIs. Extensive back-testing and simulated market conditions show that GWise outperforms traditional strategies in risk-adjusted returns while offering improved interpretability and service robustness. Our work illustrates how graph-based multiagent coordination and generative reasoning can advance real-time financial analytics as a service.

The rapid growth of multimedia content has driven the development of recommender systems. Most previous work focuses on uncovering latent relationships among items to learn better representations. However, this approach does not sufficiently account for user affinities, potentially leading to an imbalance in the structure modeling of users and items. Moreover, the sparsity and imbalance of user-item interactions further hinder effective representation learning. To address these challenges, we propose a framework called BLAST, which balances structures and interactions via large language models and optimal transport for multimodal recommendation. Specifically, we utilize large language models to summarize side information and generate user profiles. Based on these profiles, we design an intra- and inter-entity structure balancing module to capture item-item and user-user relationships, integrating these affinities into the final representations. Furthermore, we impose constraints on negative sample selection, augment the training data with false negative items and the optimal transport algorithm, thereby leading to smoother interactions. We evaluate BLAST on three real-world datasets, and the results demonstrate that our method significantly outperforms state-of-the-art baselines, which validates the superiority and effectiveness of BLAST.

Recommender systems play a critical role in many applications by providing personalized recommendations based on user interactions. However, it remains a major challenge to capture complex sequential patterns and address noise in user interaction data. While advanced neural networks have enhanced sequential recommendation by modeling high-order item dependencies, they typically assume that the noisy interaction data as the user's preferred preferences. This assumption can lead to suboptimal recommendation results. We propose a Variational Graph Auto-Encoder driven Graph Enhancement (VGAE-GE) method for robust augmentation in sequential recommendation. Specifically, our method first constructs an item transition graph to capture higher-order interactions and employs a Variational Graph Auto-Encoder (VGAE) to generate latent variable distributions. By utilizing these latent variable distributions for graph reconstruction, we can improve the item representation. Next, we use a Graph Convolutional Network (GCN) to transform these latent variables into embeddings and infer more robust user representations from the updated item embeddings. Finally, we obtain the reconstructed user check-in data, and then use a Mamba-based recommender to make the recommendation process more efficient and the recommendation results more accurate. Extensive experiments on five public datasets demonstrate that our VGAE-GE model improves recommendation performance and robustness.

With the rapid urbanization in the modern era, smart traffic profiling based on multimodal sources of data has been playing a significant role in ensuring safe travel, reducing traffic congestion and optimizing urban mobility. Most existing methods for traffic profiling on the road level usually utilize single-modality data, i.e., they mainly focus on image processing with deep vision models or auxiliary analysis on the textual data. However, the joint modeling and multimodal fusion of the textual and visual modalities have been rarely studied in road traffic profiling, which largely hinders the accurate prediction or classification of traffic conditions. To address this issue, we propose a novel multimodal learning and fusion framework for road traffic profiling, named TraffiCFUS. Specifically, given the traffic images, our TraffiCFUS framework first introduces Vision Language Models (VLMs) to generate text and then creates tailored prompt instructions for refining this text according to the specific scene requirements of road traffic profiling. Next, we apply the discrete Fourier transform to convert multimodal data from the spatial domain to the frequency domain and perform a cross-modal spectrum transform to filter out irrelevant information for traffic profiling. Furthermore, the processed spatial multimodal data is combined to generate fusion loss and interaction loss with contrastive learning. Finally, extensive experiments on four real-world datasets illustrate superior performance compared with the state-of-the-art approaches.

Graph-level anomaly detection (GLAD) identifies graphs exhibiting abnormal properties within a graph dataset.Despite promising results in this task, the state-of-the-art methods cannot be fully trusted and deployed in realistic scenarios due to their black-box nature.To alleviate this, existing methods try to explain predictions by extracting important subgraphs from each graph, as instancelevel explanations.However, instance-level explanations across all samples are costly to verify and insufficient to capture the model's general behaviors.Thus, we propose a global interpretable Graph-Level Anomaly Detection model via Prototype (GLADPro), which provides global-level explanations throughout the entire dataset, that is, the significant subgraph patterns that consistently influence the model's decisions.Specifically, GLADPro incorporates prototype learning with the information bottleneck principle, enabling prototypes to capture the most significant subgraph patterns as global-level explanations through persistent interactions with key subgraphs from input graphs.In addition, a regularization term is proposed to prevent the collapse traps with theoretical proof.Finally, we filter redundant prototypes using the maximum mean discrepancy metric.Extensive experiments demonstrate the superiority of GLADPro in anomaly detection and explainability; for instance, on the mutagen dataset, it reduces the number of explanations to verify from 1403 to only 6.

Movie Dubbing aims to convert scripts into speeches that align with the given movie clip in both temporal and emotional aspects while preserving the vocal timbre of a given brief reference audio. Existing methods focus primarily on reducing the word error rate while ignoring the importance of lip-sync and acoustic quality. To address these issues, we propose a novel dubbing architecture based on Large Language Model (LLM) and Conditional Flow Matching (CFM), named FlowDubber, which achieves high-quality audio-visual sync and pronunciation by incorporating a large speech language model with dual contrastive alignment while improving acoustic quality via Flow-based Voice Enhancing (FVE). First, we introduce Qwen2.5 as the backbone of large speech language model to learn the in-context sequence from movie scripts and reference audio. Second, the proposed semantic-aware learning focuses on capturing LLM semantic knowledge at the phoneme level, which facilitates mutual alignment with lip movement from silent video via Dual Contrastive Alignment (DCA). Third, the FVE introduces an LLM-based acoustics flow matching guidance to strengthen clarity by decoupling Classifier-Free Guidance (CFG) enhancement. Extensive experiments demonstrate that our method outperforms several state-of-the-art methods on two primary benchmarks. The demos are available at https://galaxycong.github.io/LLM-Flow-Dubber/.

Recent advancements in generative models have positioned them as one of the principal tools for sequential recommendation due to their exceptional sample diversity and generalization capabilities. Among these, diffusion model-based sequential recommenders have achieved remarkable success. However, most existing approaches still face critical challenges, resulting in suboptimal generation quality: (1) They fail to leverage multimodal knowledge for constructing item representations with well-structured distributional characteristics and semantically enriched information; (2) They predominantly rely on discrete diffusion processes, leading to high error accumulation, reduced time efficiency, and constrained controllability in generative sampling. To mitigate these challenges, we propose LSGM4Rec, a novel framework that integrates Large Language Models (LLMs) with advanced multimodal encoding models to establish multimodal fusion embeddings for items. This design ensures distinct distributional characteristics while enabling the incorporation of semantically rich modal features into guidance condition. Furthermore, we pioneer the stochastic differential equations (SDEs) for recommendation, facilitating smooth transitions between data distributions and enabling optimal trade-off between sampling efficiency and generation quality. Extensive experiments on three datasets demonstrate that LSGM4Rec outperforms existing state-of-the-art sequential recommendation methods.

Open-world object counting leverages the robust text-image alignment of pre-trained vision-language models (VLMs) to enable counting of arbitrary categories in images specified by textual queries. However, widely adopted naive fine-tuning strategies concentrate exclusively on text-image consistency for categories contained in, which leads to limited generalizability for unseen categories. In this work, we propose a plug-and-play Semantic-Driven Visual Prompt Tuning framework (SDVPT) that transfers knowledge from the training set to unseen categories with minimal overhead in parameters and inference time. First, we introduce a two-stage visual prompt learning strategy composed of Category-Specific Prompt Initialization (CSPI) and Topology-Guided Prompt Refinement (TGPR). The CSPI generates category-specific visual prompts, and then TGPR distills latent structural patterns from the VLM's text encoder to refine these prompts. During inference, we dynamically synthesize the visual prompts for unseen categories based on the semantic correlation between unseen and training categories, facilitating robust text-image alignment for unseen categories. Extensive experiments integrating SDVPT with all available open-world object counting models demonstrate its effectiveness and adaptability across three widely used datasets: FSC-147, CARPK, and PUCPR+. Code is available https://github.com/Eamon-0v0/SDVPT

Next Point-of-Interest (POI) recommendation has become a crucial task in Location-Based Social Networks (LBSNs), which provide personalized recommendations by predicting the user's next check-in locations. Commonly used models including Recurrent Neural Networks (RNNs) and Graph Convolutional Networks (GCNs) have been widely explored. However, these models face significant challenges, including the difficulty of capturing the hierarchical and tree-like structure of POIs in Euclidean space and the sparsity problem inherent in POI recommendations. To address these challenges, we propose a Hyperbolic Variational Graph Auto-Encoder (HVGAE) for next POI recommendation. Specifically, we utilize a Hyperbolic Graph Convolutional Network (Hyperbolic GCN) to model hierarchical structures and tree-like relationships by converting node embeddings from euclidean space to hyperbolic space. Then we use Variational Graph Auto-Encoder (VGAE) to convert node embeddings to probabilistic distributions, enhancing the capture of deeper latent features and providing a more robust model structure. Furthermore, we combine the Mamba4Rec recommender and Rotary Position Embedding (RoPE) and propose Rotary Position Mamba (RPMamba) to effectively utilize POI embeddings rich in sequential information, which improves the accuracy of the next POI recommendation. Extensive experiments on three public datasets demonstrate the superior performance of the HVGAE model.

Graph Neural Networks (GNNs) have emerged as powerful tools for recommendation systems, addressing challenges such as complex user-item relationships and dynamic behaviors. This paper provides a concise review of GNN applications, focusing on embedding techniques and state-of-the-art algorithms. We evaluate three GNN architectures: Homogeneous GNN, Heterogeneous GNN, and a Heterogeneous GNN enhanced with Skip-Gram node embeddings. These models are assessed on subsets of the Amazon 2023 dataset, covering Fashion, Beauty, and Musical Instruments, using metrics such as Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). Our findings highlight the superior performance of GNNs in capturing nuanced user-item interactions, improving recommendation accuracy, scalability, and adaptability. The integration of Skip-Gram embeddings further enhances item similarity modeling, enabling more personalized recommendations. We also analyze the computational efficiency of these models, offering insights for their deployment in large-scale systems. This study bridges the gap between theoretical advancements and practical applications of GNNs in recommendation systems. By synthesizing recent trends, identifying research gaps, and presenting actionable insights, it serves as a foundational reference for researchers and practitioners aiming to optimize GNN-based models for diverse scenarios. The results underscore the transformative potential of GNNs in delivering accurate, scalable, and real-time recommendations, setting the stage for future innovations in this domain.

Transformers-based trackers offer significant potential for integrating semantic interdependence between template and search features in tracking tasks. Transformers possess inherent capabilities for processing long sequences and extracting correlations within them. Several researchers have explored the feasibility of incorporating Transformers to model continuously changing search areas in tracking tasks. However, their approach has substantially increased the computational cost of an already resource-intensive Transformer. Additionally, existing Transformers-based trackers rely solely on mechanically employing multi-head attention to obtain representations in different subspaces, without any inherent bias. To address these challenges, we propose HEART (Historical Information Embedding And Subspace Re-weighting Tracker). Our method embeds historical information into the queries in a lightweight and Markovian manner to extract discriminative attention maps for robust tracking. Furthermore, we develop a multi-head attention distribution mechanism to retrieve the most promising subspace weights for tracking tasks. HEART has demonstrated its effectiveness on five datasets, including OTB-100, LaSOT, UAV123, TrackingNet, and GOT-10k.

The widespread diffusion of fake news has become a critical problem on dynamic social media worldwide, which requires effective strategies for fake news detection to alleviate its hazardous consequences for society. However, most recent efforts only focus on the features of news content and social context without realizing the benefits of large language models (LLMs) and multiple knowledge graphs (KGs), thus failing to improve detection capabilities further. To tackle this issue, we present a multiknowledge and LLM-inspired heterogeneous graph neural network for fake news detection (MiLk-FD), by combining KGs, LLMs, and graph neural networks (GNNs). Specifically, we first model news content as a heterogeneous graph (HG) containing news, entity, and topic nodes and then fuse the knowledge from three KGs to augment the factual basis of news articles. Meanwhile, we leverage TransE to initialize the knowledge features and employ LLaMa2-7B to obtain the initial feature vectors of news articles. After that, we utilize the devised HG transformer to learn news embeddings with specific feature distribution in high-dimensional spaces by aggregating neighborhood information according to metapaths. Finally, a classifier based on multilayer perceptron (MLP) is trained to predict each news article as fake or true. Through experiments, we demonstrate that our proposed framework surpasses ten baselines according to accuracy, precision, F1-score, recall, and ROC in four public real-world benchmarks (i.e., COVID-19, FakeNewsNet, PAN2020, Liar).

As generative Artificial Intelligence (AI) technologies evolve, they offer unprecedented potential to automate and enhance various tasks, including coding. Natural Language-Oriented Programming (NLOP), a vision introduced in this paper, harnesses this potential by allowing developers to ar-ticulate software requirements and logic in their natural language, thereby democratizing software creation. This approach streamlines the development process and significantly lowers the barrier to entry for software engineering, making it feasible for non-experts to contribute effectively to software projects. By simplifying the transition from concept to code, NLOP can accelerate development cycles, enhance collaborative efforts, and reduce misunderstandings in requirement specifications. This paper reviews various programming models, assesses their contri-butions and limitations, and highlights that natural language will be the new programming language. Through this comparison, we illustrate how NLOP stands to transform the landscape of software engineering by fostering greater inclusivity and innovation.

Federated Learning (FL) has emerged as a promising solution in Edge Computing (EC) environments to process the proliferation of data generated by edge devices. By collaboratively optimizing the global machine learning models on distributed edge devices, FL circumvents the need for transmitting raw data and enhances user privacy. Despite practical successes, FL still confronts significant challenges including constrained edge device resources, multiple tasks deployment, and data heterogeneity. However, existing studies focus on mitigating the FL training costs of each single task whereas neglecting the resource consumption across multiple tasks in heterogeneous FL scenarios. In this paper, we propose Heterogeneous Federated Learning with Local Parameter Sharing (FedLPS) to fill this gap. FedLPS leverages principles from transfer learning to facilitate the deployment of multiple tasks on a single device by dividing the local model into a shareable encoder and task-specific encoders. To further reduce resource consumption, a channel-wise model pruning algorithm that shrinks the footprint of local models while accounting for both data and system heterogeneity is employed in FedLPS. Additionally, a novel heterogeneous model aggregation algorithm is proposed to aggregate the heterogeneous predictors in FedLPS. We implemented the proposed FedLPS on a real FL platform and compared it with state-of-the-art (SOTA) FL frameworks. The experimental results on five popular datasets and two modern DNN models illustrate that the proposed FedLPS significantly outperforms the SOTA FL frameworks by up to 4.88% and reduces the computational resource consumption by 21.3%. Our code is available at: https://github.com/jyzgh/FedLPS.

Deep learning is increasingly crucial in scientific discovery, accelerating research in various fields. Exploring brain science using deep learning has garnered significant interest, particularly in the recognition of brain disorders. However, existing methods face limitations in the discriminability of learned brain graph representations and the identification of neurological biomarkers associated with brain disorders. Moreover, the lack of model explainability leads to suboptimal brain graph analysis. To address these challenges, we propose a bi-level brain graph contrastive learning framework with an interpretable learning kernel for brain disorder analysis. Our framework diverges from traditional graph contrastive learning methods by augmenting meaningful brain graphs using the counterfactual thinking-based mechanism, ensuring reliable graph generation that preserves discriminative information. Secondly, we introduce bi-level contrastive loss with intra-and inter-class contrasts to enhance the brain graph representation learning. Most importantly, we design an interpretable brain graph learning kernel to highlight critical regions and connections, thereby facilitating the discovery of potential neurological biomarkers associated with brain disorders in subsequent analysis. The effectiveness of our method11https://githuh.com/JustinGie/Cf-BCL in brain graph representation learning and discriminative substructure detection is demonstrated through the evaluation of disorder prediction and pathogenic analysis on three real-world brain disorder datasets. Moreover, our framework may provide novel insights into brain science based on the results of comparison with medical research.

Data activities and technologies in modern organisations are becoming more diverse and complex, and continuously operate, change, and evolve as an enterprise data ecosystem (EDE). This fast-changing environment presents both challenges and opportunities for many computing and design concepts and approaches and raises the questions whether and how they can be continuously and effectively applied. This paper analyses operation features and complexity of EDE and delves into design considerations for data products and Data-as-a-Service (DaaS) to play their roles as solutions for data practice management improvement, integration, interoperability and design of services in data space.

Big Data has various formats, substandard data models, and diverse semantics. Understanding this vast, continuous, and constantly evolving data is challenging and requires sophisticated data science pipelines that not only turn the raw data into contextual knowledge and insights but also generate meaningful and personalized narratives to facilitate the understanding and interpretation of the Big Data. Existing methods often fail to provide the necessary context and personalization to make data actionable to diverse stakeholders. To achieve this goal, we introduce the Open Story Model (OSM), a novel data science pipeline that: i) transforms the unprocessed data into a structured Knowledge Graph based on a unified graph data model of typed nodes and edges; ii) generates meaningful narratives on top of the Knowledge Graph; and (iii) link narratives to construct personalized stories for various stakeholders based on their goals. We model each story as an interactive data product to allow stakeholders to engage with and manipulate data in real-time to explore, analyze, and visualize the information in ways that suit their specific needs and questions. To demonstrate the practical application and efficacy of the OSM framework, we focus on a use case in the health domain and explore how OSM facilitates the detection and prediction of cancer by transforming the data into personalized stories for various stakeholders.

Federated learning addresses privacy concerns in multimedia recommender systems by enabling collaborative model training without exchanging raw data. However, existing federated recommendation models are mainly based on basic backbones like Matrix Factorization (MF), which are inadequate to capture complex implicit interactions between users and multimedia content. Graph Convolutional Networks (GCNs) offer a promising method by utilizing the information from high-order neighbors, but face challenges in federated settings due to problems such as over-smoothing, data heterogeneity, and elevated communication expenses. To resolve these problems, we propose a Cluster-driven Personalized Federated Recommender System with Interest-aware Graph Convolution Network (CPF-GCN) for multimedia recommendation. CPF-GCN comprises a local interest-aware GCN module that optimizes node representations through subgraph-enhanced adaptive graph convolution operations, mitigating the over-smoothing problem by adaptively extracting information from layers and selectively utilizing high-order connectivity based on user interests. Simultaneously, a cluster-driven aggregation approach at the server significantly reduces communication costs by selectively aggregating models from clusters. The aggregation produces a global model and cluster-level models, combining them with the user's local model allows us to tailor the recommendation model for the user, achieving personalized recommendations. Moreover, we propose an adversarial optimization technique to further augment the robustness of CPF-GCN. Experiments on three datasets demonstrate that CPF-GCN significantly outperforms the state-of-the-art models.

Federated learning (FL) has emerged as a prominent machine learning paradigm in edge computing environments, enabling edge devices to collaboratively optimize a global model without sharing their private data. However, existing FL frameworks suffer from efficacy deterioration due to the system heterogeneity inherent in edge computing, especially in the presence of domain shifts across local data. In this paper, we propose a heterogeneous FL framework DapperFL, to enhance model performance across multiple domains. In DapperFL, we introduce a dedicated Model Fusion Pruning (MFP) module to produce personalized compact local models for clients to address the system heterogeneity challenges. The MFP module prunes local models with fused knowledge obtained from both local and remaining domains, ensuring robustness to domain shifts. Additionally, we design a Domain Adaptive Regularization (DAR) module to further improve the overall performance of DapperFL. The DAR module employs regularization generated by the pruned model, aiming to learn robust representations across domains. Furthermore, we introduce a specific aggregation algorithm for aggregating heterogeneous local models with tailored architectures and weights. We implement DapperFL on a real-world FL platform with heterogeneous clients. Experimental results on benchmark datasets with multiple domains demonstrate that DapperFL outperforms several state-of-the-art FL frameworks by up to 2.28%, while significantly achieving model volume reductions ranging from 20% to 80%. Our code is available at: https://github.com/jyzgh/DapperFL.

Explainable Artificial Intelligence (XAI) aims to make artificial intelligence (AI) systems transparent and understandable to humans, providing clear explanations for the decisions made by AI models. This paper presents a novel pipeline and a digital dashboard that provides a user-friendly platform for interpreting the results of machine learning algorithms using XAI technology. The dashboard utilizes evidence-based design principles to deliver information clearly and concisely, enabling users to better understand the decisions made by their algorithms. We integrate XAI services into the dashboard to explain the algorithm's predictions, allowing users to understand how their models function and make informed decisions. We demonstrate a motivating scenario in banking and present how the proposed system enhances transparency and accountability and improves trust in the technology.

In this paper, we propose an interpretable brain graph contrastive learning framework, which aims to learn brain graph representations by a unsupervised way for disorder prediction and pathogenic analysis. Our framework consists of two key designs: We first utilize the controllable data augmentation strategy to perturb unimportant structures and attribute features for the generation of brain graphs. Then, considering that the difference of healthy and patient brain graphs is small, we introduce hard negative sample evaluation to weight negative samples of the contrastive loss, which can learn more discriminative brain graph representations. More importantly, our method can observe salient brain regions and connections for pathogenic analysis. We conduct disorder prediction and interpretable analysis experiments on three real-world neuroimaging datasets to demonstrate the effectiveness of our framework.

Spatially resolved transcriptomics (SRT) is a pioneering method for simultaneously studying morphological contexts and gene expression at single-cell precision. Data emerging from SRT are multifaceted, presenting researchers with intricate gene expression matrices, precise spatial details and comprehensive histology visuals. Such rich and intricate datasets, unfortunately, render many conventional methods like traditional machine learning and statistical models ineffective. The unique challenges posed by the specialized nature of SRT data have led the scientific community to explore more sophisticated analytical avenues. Recent trends indicate an increasing reliance on deep learning algorithms, especially in areas such as spatial clustering, identification of spatially variable genes and data alignment tasks. In this manuscript, we provide a rigorous critique of these advanced deep learning methodologies, probing into their merits, limitations and avenues for further refinement. Our in-depth analysis underscores that while the recent innovations in deep learning tailored for SRT have been promising, there remains a substantial potential for enhancement. A crucial area that demands attention is the development of models that can incorporate intricate biological nuances, such as phylogeny-aware processing or in-depth analysis of minuscule histology image segments. Furthermore, addressing challenges like the elimination of batch effects, perfecting data normalization techniques and countering the overdispersion and zero inflation patterns seen in gene expression is pivotal. To support the broader scientific community in their SRT endeavors, we have meticulously assembled a comprehensive directory of readily accessible SRT databases, hoping to serve as a foundation for future research initiatives.

Livestock, throughout their lifespan, are transported to multiple destinations before being processed into consumable goods. The assurance of authentic product delivery hinges on the presence of a reliable, intelligent identification system. However, extant livestock identification methodologies, primarily relying on radio frequency identification (RFID) ear tags, are vulnerable to loss, failure, and cases of misidentification or improper substitution. This paper introduces an Artificial Intelligence (AI)-enabled system to rectify these issues by leveraging deep learning facial recognition for cattle re-identification. It utilises an integrated approach combining the Only Look Once version 5 (YOLOv5) algorithm for cattle face detection and the Siamese Neural Network (SNN) for subsequent recognition. The system was rigorously tested on a prepared dataset consisting of 2,500 cattle face images, demonstrating an impressive accuracy of 95.13% when supplied with a single query image and a 20-image sample per cow from our dataset. This system can be deployed across diverse environments, including farms, cargo areas, and sale yards, without necessitating model re-training. Furthermore, it can be fine-tuned to identify other farm animals, indicating its broad applicability.

Automatic image annotation (AIA) is a fundamental and challenging task in computer vision. Considering the correlations between tags can lead to more accurate image understanding, benefiting various applications, including image retrieval and visual search. While many attempts have been made to incorporate tag correlations in annotation models, the method of constructing a knowledge graph based on external knowledge sources and hyperbolic space has not been explored. In this paper, we create an attributed knowledge graph based on vocabulary, integrate external knowledge sources such as WordNet, and utilize hyperbolic word embeddings for the tag representations. These embeddings provide a sophisticated tag representation that captures hierarchical and complex correlations more effectively, enhancing the image annotation results. In addition, leveraging external knowledge sources enhances contextuality and significantly enriches existing AIA datasets. We exploit two deep learning-based models, the Relational Graph Convolutional Network (R-GCN) and the Vision Transformer (ViT), to extract the input features. We apply two R-GCN operations to obtain word descriptors and fuse them with the extracted visual features. We evaluate the proposed approach using three public benchmark datasets. Our experimental results demonstrate that the proposed architecture achieves state-of-the-art performance across most metrics on Corel5k, ESP Game, and IAPRTC-12. • Knowledge graph with hyperbolic embeddings enhances automatic image annotation tags. • External knowledge sources, like WordNet, enrich the annotation datasets. • R-GCN and ViT models effectively fuse textual and visual features. • Calibrated predictions refine class probabilities for more reliable annotations.

Significant progress has been made in video captioning in recent years. However, most existing methods directly learn from all given captions without distinguishing the styles of captions. The large diversity in these captions might bring ambiguity to the model learning. To address this issue, we propose a style-aware two-stage learning framework. In the first stage, the model is trained with captions of separate styles, including length style (short, medium, long), action style (single action or multiple actions), and object style (one object or more). For efficiency, a shared model with multiple individual style vectors is learned. In the second stage, a video style encoder is devised to capture style information from the input video, and it outputs a guidance signal of how to utilize the style vectors for the final caption generation. Without whistles and bells, our method achieves state-of-the-art performance on three widely-used public datasets, MSVD, MSR-VTT and VATEX. The source code and trained models will be made available to the public.

Guest Editorial: Special Issue on Generating Human Readable Explanations in NLP

Event extraction (EE) is a crucial research task for promptly apprehending event information from massive textual data. With the rapid development of deep learning, EE based on deep learning technology has become a research hotspot. Numerous methods, datasets, and evaluation metrics have been proposed in the literature, raising the need for a comprehensive and updated survey. This article fills the research gap by reviewing the state-of-the-art approaches, especially focusing on the general domain EE based on deep learning models. We introduce a new literature classification of current general domain EE research according to the task definition. Afterward, we summarize the paradigm and models of EE approaches, and then discuss each of them in detail. As an important aspect, we summarize the benchmarks that support tests of predictions and evaluation metrics. A comprehensive comparison among different approaches is also provided in this survey. Finally, we conclude by summarizing future research directions facing the research area.

Automatic detection of Cyberbullying is a challenging task due to the availability of limited trained data, which is usually noisy and inherently multimodal. Transfer learning over pre-trained BERT-based language models has succeeded in various complex use cases like sequence-to-sequence translation and text classification. These methods mainly utilize transformer models to learn the word and sentence-level relationships. While they have demonstrated promising results, they only focus on textual features without taking contextual and structural information into account. Moreover, due to the data-heavy nature of BERT-based models, they may fail to model all the desired relationships if not adequate training data is provided to them during the fine-tuning process. In this paper, we propose a novel Session-level Contextualized Transformer-based architecture for Cyberbullying Detection (SECTR-CD), which can leverage transfer learning for modeling word-level attention while also being able to model sentence-level relationships in large bodies of text. The model is also capable of utilizing other contextual features from various modalities like images and social information. Our experimental results indicate remarkable improvement in the Cyberbullying detection task even in the presence of limited training samples.

Text summarization aims to condense text documents into a concise textual summary, which improves the efficiency of people in comprehending information. While deep learning-based summarization methods for individual documents have achieved good performance, there is an increasing demand for summarizing multiple related documents of a topic or event can yield a more coherent and succinct summary of the document set. However, the characteristics of multiple documents with more information, longer texts, and different styles impose new challenges to existing methods in dealing with the multi-aspect of a topic or an event. Therefore, in this paper, we propose a novel multi-granularity model with entity recognition for better sentence ranking and capturing the key information of different documents with a comprehensive and accurate summary. Specifically, we use PRIMERA as a token encoder based on the encoder-decoder framework. Then, a named entity recognition model is trained to identify key elements in documents such as people, location, organization, etc. The proposed model will focus more on these key elements. Based on the named entity recognition results, we further devise a sentence ranking module that allows the model to assign different weights to different sentences based on the sum of the frequencies of the entities contained in the sentences. Finally, based on the results of the multi-granularity encoding vector, the decoder can generate a comprehensive and accurate summary. To evaluate the performance of our proposed model, we conducted experiments on CoNLL2003, DUC2003, and DUC2004, which demonstrated the performance improvement of our proposed method over four previous models.

Anomaly detection is one of the crucial research topics in artificial intelligence, encompassing various fields such as health monitoring, network intrusion detection, and fraud detection in financial transactions. Deep anomaly detection (DAD) methods are considered as the effective approaches for addressing complex anomaly detection problems. Among them, the deep isolation forest methods have gained rapid development recently due to their simplicity in parameter turning and efficiency in model training. The existing deep isolation forest approaches are all based on representation learning, while OptiForest theoretically proves the crucial role of the tree structure in isolation forest based methods. In this paper, we analyse the search space of isolation trees under specific data instances and address the challenges in finding optimal isolation forest. Based on the theoretical underpinning and genetic algorithm, we design a deep model DOIForest with two mutation schemes and solution selection, which learns the optimal isolation forest and optimises the parameters in data partitioning. Extensive experiments on both synthetic dataset and a series of real-world datasets demonstrate that our approach can achieve better detection accuracy and robustness than the state-of-the-arts.

With the development of modern technology, the application of various types of devices in life has become more extensive, especially with the emergence of the Internet of Things (IoT), which makes a large number of devices to be connected to the network. However, while bringing convenience to life, attacks against IoT devices have also begun to appear and become one of the most concerned issues. These IoT devices produce and transmit data containing important user information every second. Attackers are targeting this characteristic to initiate further malicious attacks on these valuable data. In order to protect users from such attacks, it is important that threats can be detected and identified in time before they cause damage. To address this issue, this paper first reviewed the current state-of-the-art anomaly detection methods and based on the finding of uncovered areas of existing methods, this paper proposed a new anomaly detection framework leveraging the combination of Federated Learning and Isolation Forest. The framework performs tree construction on the clients’ end and further uploads the encrypted data containing the nodes’ information of the trees to the central server for the forest construction, after multiple interactions, the abnormal behaviour in the clients could be able to be identified more effectively and ultimately improved the accuracy of the detection results while protecting the privacy of the client data.

Intelligence refers to the ability to acquire and apply knowledge and skills, which comprises three fundamental components, namely knowledge, experience, and creativity. Consequently, there exist three primary Artificial Intelligence (AI) systems, namely Analytical AI, Cognitive AI, and Generative AI. Analytical AI is primarily concerned with comprehending the data and transforming it into contextualized data and knowledge. On the other hand, Cognitive AI is centered on understanding experience and aims to annotate, enrich, and utilize the knowledge, to facilitate decision-making. Lastly, Generative AI delves into the neural mechanisms involved in creative thinking and problem-solving, with a focus on enhancing the process of acquiring and applying knowledge and skills. This paper presents Knowledge Base 4.0 as the backend data for AI engines, which allows for linking knowledge and experience to enable empowering generative AI. The objective is not only to facilitate generating new content (such as text and images) but also to generate new processes when/if needed. We present the architecture of Knowledge Base 4.0 and the design and development of data services that construct and maintain this robust Knowledge Base. Additionally, we provide use cases in various domains, including health, policing, banking, and education.

Generative Pre-trained Transformer (GPT) is a state-of-the-art machine learning model capable of generating human-like text through natural language processing (NLP). GPT is trained on massive amounts of text data and uses deep learning techniques to learn patterns and relationships within the data, enabling it to generate coherent and contextually appropriate text. This position paper proposes using GPT technology to generate new process models when/if needed. We introduce ProcessGPT as a new technology that has the potential to enhance decision-making in data-centric and knowledge-intensive processes. ProcessGPT can be designed by training a generative pre-trained transformer model on a large dataset of business process data. This model can then be fine-tuned on specific process domains and trained to generate process flows and make decisions based on context and user input. The model can be integrated with NLP and machine learning techniques to provide insights and recommendations for process improvement. Furthermore, the model can automate repetitive tasks and improve process efficiency while enabling knowledge workers to communicate analysis findings, support evidence, and make decisions. ProcessGPT can revolutionize business process management (BPM) by offering a powerful tool for process automation and improvement. Finally, we demonstrate how ProcessGPT can be a powerful tool for augmenting data engineers in maintaining data ecosystem processes within large bank organizations. Our scenario highlights the potential of this approach to improve efficiency, reduce costs, and enhance the quality of business operations through the automation of data-centric and knowledge-intensive processes. These results underscore the promise of ProcessGPT as a transformative technology for organizations looking to improve their process workflows.

The exponential growth of textual data has created a crucial need for tools that assist users in extracting meaningful insights. Traditional document summarization approaches often fail to meet individual user requirements and lack structure for efficient information processing. To address these limitations, we propose Summation, a hierarchical personalized concept-based summarization approach. It synthesizes documents into a concise hierarchical concept map and actively engages users by learning and adapting to their preferences. Using a Reinforcement Learning algorithm, Summation generates personalized summaries for unseen documents on specific topics. This framework enhances comprehension, enables effective navigation, and empowers users to extract meaningful insights from large document collections aligned with their unique requirements.

Anomaly detection plays an increasingly important role in various fields for critical tasks such as intrusion detection in cybersecurity, financial risk detection, and human health monitoring. A variety of anomaly detection methods have been proposed, and a category based on the isolation forest mechanism stands out due to its simplicity, effectiveness, and efficiency, e.g., iForest is often employed as a state-of-the-art detector for real deployment. While the majority of isolation forests use the binary structure, a framework LSHiForest has demonstrated that the multi-fork isolation tree structure can lead to better detection performance. However, there is no theoretical work answering the fundamentally and practically important question on the optimal tree structure for an isolation forest with respect to the branching factor. In this paper, we establish a theory on isolation efficiency to answer the question and determine the optimal branching factor for an isolation tree. Based on the theoretical underpinning, we design a practical optimal isolation forest OptIForest incorporating clustering based learning to hash which enables more information to be learned from data for better isolation quality. The rationale of our approach relies on a better bias-variance trade-off achieved by bias reduction in OptIForest. Extensive experiments on a series of benchmarking datasets for comparative and ablation studies demonstrate that our approach can efficiently and robustly achieve better detection performance in general than the state-of-the-arts including the deep learning based methods.

The goal of crime prediction is to forecast the number of crime incidents at each region of a city based on the historical crime data. It has attracted a great deal of attention from both academic and industrial communities due to its considerable significance in improving urban safety and reducing financial losses. Although much progress has been made in this field, most of the existing approaches assume that the historical crime data are complete, which does not hold in many real-world scenarios. Meanwhile, crime incidents are affected by multiple factors and have intricate spatial, temporal, and categorical correlations, which are not fully utilized by the current methods. In this article, we propose a novel tensor decomposition based framework, named TD-Crime, to conduct prediction directly on the incomplete crime data. Specifically, we first organize the crime data as a tensor and then apply the nonnegative CP decomposition to it, which not only provides a natural solution to the missing data problem but also captures the spatial, temporal, and categorical correlations implicitly. Moreover, we attempt to exploit the spatial and temporal correlations explicitly by directly learning from the crime data to further improve the forecasting performance. Finally, we obtain a joint optimization problem and present an efficient alternating optimization scheme to find a satisfactory solution. Extensive experiments on the real-world crime datasets show that TD-Crime can address the crime prediction task effectively under different missing data scenarios.

The rapid development of the Internet of Vehicles has substantially boosted the prevalence of vehicular social networks (VSN). However, content security has gradually been a latent threat to the stable operation of VSN. The VSN is a time-varying environment and mixed with various real or fake contents, which brings great challenges to the sustainability of VSN. To establish a sustainable VSN, it is of practical value to possess a strong ability for fake content detection. Related works can be divided into the global semantics-based approaches and the local semantics-based approaches, though both with limitations. Leveraging these two different approaches, this paper proposes a fake content detection model based on the mixed graph neural networks (GNN) for sustainable VSN. It takes GNN as the bottom architecture and integrates both convolution neural networks and recurrent neural networks to capture two aspects of semantics. Such a mixed detection framework is expected to possess a better detection effect. A number of experiments were conducted on two social network datasets for evaluation, and the results indicated that the detection effect can be improved by about 5%-15% compared with baseline methods.

The proliferation of the Internet of Things (IoT) has led to the design and incorporation of innovative user control mechanisms, one category based on brain-derived biometric data and known as Brain Control Interface (BCI). BCI devices measure brain signals in EEG and allow users to interact with computerised systems, such as the Industrial Internet of Things (IIoT), intuitively. However, the utilisation of EEG data in the IIoT for actuator control and the collection of such biometric data as evidence for policing the IIoT introduces considerable implications for the users’ cognitive privacy. Thus, considering the importance of cognitive privacy in an evolving era of smart environments, it becomes imperative to develop methods that can ensure the cognitive privacy of users. Cognitive privacy also protects the anonymity of corporations and law enforcement. Furthermore, it maintains the inherent information found in EEG measurements, used by various machine and deep learning models. This paper proposes a novel deep learning-based human cognitive privacy framework, named DeepCog, that ensures users’ privacy through the application of feature transforming normalisation. A deep MLP model then processes the encoded data to classify samples according to an integer-based subject ID, enabling the framework to select the correct secondary deep MLP model (one for each subject) to identify eye activity. Our experiments indicate high accuracy of 93.4%, with precision 93.2% and recall 93.8%, outperforming compelling techniques.

Artificial intelligence (AI) enables machines to learn from human experience, adjust to new inputs, and perform intelligent tasks without human intervention. AI is progressing rapidly and is transforming the way businesses operate, from process automation to cognitive augmentation of tasks and intelligent process/data analytics. However, the main challenge for the AI system users is to comprehend and trust the result of AI algorithms and methods. To address this challenge, we first study the recent techniques in the area of eXplainable Artificial Intelligence (XAI). Then, we introduce a novel XAI process to facilitate producing explainable models while maintaining a high level of learning performance. We present an interactive evidence-based approach to assist the users in comprehending and trusting the results and outputs generated by AI-enabled algorithms, resulting in developing a digital dashboard to facilitate inter-acting with the algorithm. Lastly, we discuss how the proposed XAI method can significantly improve the confidence of data scientists in understanding the result of AI-enabled algorithms with an application in the banking domain for analyzing customer transactions.

Mental health disorders may cause severe consequences for countries’ economies and health. Identifying early signs of these disorders is vital. The state-of-the-art research in identifying mental health disorder patterns from textual data, uses hand-labeled training sets, especially when a domain expert’s knowledge is required to analyze various symptoms in a patient. This task could be time-consuming and expensive. To address this challenge, in this paper, we study and analyze the various clinical and non-clinical approaches to identifying mental health disorders. We leverage the domain knowledge and expertise in cognitive science to build a domain-specific Knowledge Base for the mental health disorder concepts and patterns. We present a weaker form of supervision by facilitating and generating training data from a domain-specific Knowledge Base. We adopt a typical scenario for analyzing social media to identify depression symptoms from the textual content generated by social users.

Graph anomaly detection in this paper aims to distinguish abnormal nodes that behave differently from the benign ones accounting for the majority of graph-structured instances. Receiving increasing attention from both academia and industry, yet existing research on this task still suffers from two critical issues when learning informative anomalous behavior from graph data. For one thing, anomalies are usually hard to capture because of their subtle abnormal behavior and the shortage of background knowledge about them, which causes severe anomalous sample scarcity. Meanwhile, the overwhelming majority of objects in real-world graphs are normal, bringing the class imbalance problem as well. To bridge the gaps, this paper devises a novel Data Augmentation-based Graph Anomaly Detection (DAGAD) framework for attributed graphs, equipped with three specially designed modules: 1) an information fusion module employing graph neural network encoders to learn representations, 2) a graph data augmentation module that fertilizes the training set with generated samples, and 3) an imbalance-tailored learning module to discriminate the distributions of the minority (anomalous) and majority (normal) classes. A series of experiments on three datasets prove that DAGAD outperforms ten state-of-the-art baseline detectors concerning various mostly-used metrics, together with an extensive ablation study validating the strength of our proposed modules.

Intelligence is the ability to learn from experience. Knowledge workers in knowledge-intensive processes develop invaluable domain-specific expertise and knowledge over time. Accordingly, it is vital for organizations to capture this knowledge (which is hidden in the biological Neural Network of subject-matter experts) and enable novices/inexperienced knowledge workers to benefit from that in choosing the best next steps. This position paper proposes linking weak supervision and crowd-sourcing techniques to incorporate knowledge in a continual fashion based on estimating uncertainty or errors from the existing knowledge and learning models. Applications may include handling cold start and concept drift situations. We discuss the design of an intelligent Knowledge Base (KB), namely KB 4.0, for mimicking the knowledge of domain experts in knowledge-intensive processes, and using this knowledge to facilitate auto labelling of the data to be used in learning algorithms. We present a motivating scenario in police investigation processes, and argue how inexperienced investigators can benefit from such a domain-specific KB in law enforcement.

With the ever-increasing popularity of microservice architecture, a considerable number of enterprises or organizations have encapsulated their complex business services into various lightweight functions as published them accessible APIs (Application Programming Interfaces). Through keyword search, a software developer could select a set of APIs from a massive number of candidates to implement the functions of a complex mashup, which reduces the development cost significantly. However, traditional keyword search methods for APIs often suffer from several critical issues such as functional compatibility and limited diversity in search results, which may lead to mashup creation failures and lower development productivity. To deal with these challenges, this paper designs DAWAR, a diversity-aware Web APIs recommendation approach that finds diversified and compatible APIs for mashup creation. Specifically, the APIs recommendation problem for mashup creating is modelled as a graph search problem that aims to find the minimal group Steiner trees in a correlation graph of APIs. DAWAR innovatively employs the determinantal point processes to diversify the recommended results. Empirical evaluation is performed on commonly-used real-world datasets, and the statistic results show that DAWAR is able to achieve significant improvements in terms of recommendation diversity, accuracy, and compatibility.

Graph anomaly detection, here, aims to find rare patterns that are significantly different from other nodes. Attributed graphs containing complex structure and attribute information are ubiquitous in our life scenarios such as bank account transaction graph and paper citation graph. Anomalous nodes on attributed graphs show great difference from others in the perspectives of structure and attributes, and give rise to various types of graph anomalies. In this paper, we investigate three types of graph anomalies: local, global, and structure anomalies. And, graph neural networks (GNNs) based anomaly detection methods attract considerable research interests due to the power of modeling attributed graphs. However, the convolution operation of GNNs aggregates neighbors information to represent nodes, which makes node representations more similar and cannot effectively distinguish between normal and anomalous nodes, thus result in sub-optimal results. To improve the performance of anomaly detection, we propose a novel community-aware attributed graph anomaly detection framework (ComGA). We design a tailored deep graph convolutional network (tGCN) to anomaly detection on attributed graphs. Extensive experiments on eight real-life graph datasets demonstrate the effectiveness of ComGA.

BACKGROUND: Colorectal cancer (CRC) is one of the leading causes of cancer-related deaths worldwide. Recent studies have observed causative mutations in susceptible genes related to colorectal cancer in 10 to 15% of the patients. This highlights the importance of identifying mutations for early detection of this cancer for more effective treatments among high risk individuals. Mutation is considered as the key point in cancer research. Many studies have performed cancer subtyping based on the type of frequently mutated genes, or the proportion of mutational processes. However, to the best of our knowledge, combination of these features has never been used together for this task. This highlights the potential to introduce better and more inclusive subtype classification approaches using wider range of related features to enable biomarker discovery and thus inform drug development for CRC. RESULTS: In this study, we develop a new pipeline based on a novel concept called 'gene-motif', which merges mutated gene information with tri-nucleotide motif of mutated sites, for colorectal cancer subtype identification. We apply our pipeline to the International Cancer Genome Consortium (ICGC) CRC samples and identify, for the first time, 3131 gene-motif combinations that are significantly mutated in 536 ICGC colorectal cancer samples. Using these features, we identify seven CRC subtypes with distinguishable phenotypes and biomarkers, including unique cancer related signaling pathways, in which for most of them targeted treatment options are currently available. Interestingly, we also identify several genes that are mutated in multiple subtypes but with unique sequence contexts. CONCLUSION: Our results highlight the importance of considering both the mutation type and mutated genes in identification of cancer subtypes and cancer biomarkers. The new CRC subtypes presented in this study demonstrates distinguished phenotypic properties which can be effectively used to develop new treatments. By knowing the genes and phenotypes associated with the subtypes, a personalized treatment plan can be developed that considers the specific phenotypes associated with their genomic lesion.

Performance analysis and optimization is a critical task for the successful development of cloud computing systems and services. Unfortunately, performance analysis and optimization remains complicated and challenging due to several unique characteristics in cloud computing such as stochastic service requests, request sequencing strategies, and request distribution methods. In this paper, we present a comprehensive survey on the performance analysis and optimization for stochastic cloud service requests. By analyzing the main entities and activities in the common routines of performance analysis, we first propose a generic performance analysis framework, which contains five fundamental characteristics: Request, Sequencing, Queue, Distribution and Services. Practical factors of each characteristic are analyzed. We discuss the effects of each characteristic of the framework on optimization objectives including cost, profit, response time, and energy consumption. We then systematically review and compare 13 representative queuing models using the proposed framework. Based on the practical factors of the five characteristics and along with the current research efforts, we also identify several research opportunities and challenges.

The objective of fraud detection is to distinguish fraudsters from normal users. In graph/network environments, both fraudsters and normal users are modeled as nodes, and the connections between those nodes are represented as edges. Fraudsters typically try to camouflage themselves with “normal” behaviors, say, by deliberately establishing many connections to normal users. Such camouflage inherently makes their appearance inconsistent with the essence of what it is to be normal, and gives rise to inconsistencies in the graph. In this paper, we investigate three aspects of these graph inconsistencies: features, topologies, and relations. To date, graph-based fraud detectors have shown a rather limited capability to fuse information about different types of inconsistencies. Apart from that, there is another problem of imbalance to overcome. This is because fraudsters usually only account for a very small percentage of all users. To achieve a promising capability, i.e., dual-resistant to graph inconsistency and imbalance, we present a new fraud detection model FRAUDRE based on Graph Neural Networks. Extensive experiments comparing eight up-to-date baselines on two real-world datasets, Amazon and YelpChi, demonstrate the superiority of FRAUDRE.

Digitization, i.e., the process of converting information into a digital format, may provide various opportunities and challenges for businesses. In this context, one of the main challenges would be to accurately classify numerous scanned documents uploaded every day by customers as usual business processes. The current study has proposed a robust fusion model to define the type of identity documents accurately. The proposed approach is based on two different methods in which images are classified based on their visual features and text features. A novel model based on statistics and regression has been proposed to calculate the confidence level for the feature-based classifier. A fuzzy-mean fusion model has been proposed to combine the classifier results based on their confidence score. The proposed approach has been implemented using Python and experimentally validated on synthetic and real-world datasets. The performance is evaluated using the Receiver Operating Characteristic (ROC) curve analysis.

Recently, financial institutes have been dealing with an increase in financial crimes. In this context, financial services firms started to improve their vigilance and use new technologies and approaches to identify and predict financial fraud and crime possibilities. This task is challenging as institutions need to upgrade their data and analytics capabilities to enable new technologies such as Artificial Intelligence (AI) to predict and detect financial crimes. In this paper, we put a step towards AI-enabled financial crime detection in general and money laundering detection in particular to address this challenge. We study and analyse the recent works done in financial crime detection and present a novel model to detect money laundering cases with minimum human intervention needs.

Influence maximization, i.e., the problem of finding a small subset of nodes in a social network which can maximize the propagation of influence, has the potential to become a vital asset to identify and predict mental health related issues such as, predicting suicide, bulling, and radicalization. For example, predictive analytics in mental health can enable analyzing and exploring the factors involved in influencing people to participate in extremist activities. To address this challenge, in this paper, we analyze the influence maximization in mental health from effectiveness, efficiency, and scalability viewpoints. We present a social data analytics pipeline to enable analysts to engage with social data to explore the potential online radicalization. According to the predictive analytics, a particle swarm optimization influence maximization algorithm is proposed to facilitate selecting potential influential nodes. A context analytics algorithm is proposed to analyze the social data and the user activity patterns to learn how influence flows in social networks. We conducted intensive experiments based on real dataset and illustrate the effectiveness and efficiency of the proposed algorithms.

The ever-increasing amount of text data makes it challenging for humans to extract the required information. This problem is even more vital for the medical domain, where accessing up-to-date information is essential. Physicians and researchers face diverse and extensive medical sources such as medical journal articles, websites, or patient records. Therefore, they require to analyze them based on their interests and needs quickly. Intelligent content summarization approaches are assisting tools in such situations to provides an overview of a set of documents. However, the summary is required to be tailored to two different users type preferences: the physician and the patient. This paper proposes a novel embedding method, called Summary2vec, where each summary is presented by a fixed -length vector covering various aspects of information space. Summary2vec is remedial to design automatic services for various analytic purposes that require information-seeking activity. We leverage Summary2vec to produce a hierarchical summarization structure to enable users navigating through the hierarchy to gain more elaborated information upon request by engaging them.

Intelligent Transportation Systems (ITS), especially Autonomous Vehicles (AVs), are vulnerable to security and safety issues that threaten the lives of the people. Unlike manual vehicles, the security of communications and computing components of AVs can be compromised using advanced hacking techniques, thus barring AVs from the effective use in our routine lives. Once manual vehicles are connected to the Internet, called the Internet of Vehicles (IoVs), it would be exploited by cyber-attacks, like denial of service, sniffing, distributed denial of service, spoofing and replay attacks. In this article, we present a deep learning-based Intrusion Detection System (IDS) for ITS, in particular, to discover suspicious network activity of In-Vehicles Networks (IVN), vehicles to vehicles (V2V) communications and vehicles to infrastructure (V2I) networks. A Deep Learning architecture-based Long-Short Term Memory (LSTM) autoencoder algorithm is designed to recognize intrusive events from the central network gateways of AVs. The proposed IDS is evaluated using two benchmark datasets, i.e., the car hacking dataset for in-vehicle communications and the UNSW-NB15 dataset for external network communications. The experimental results demonstrated that our proposed system achieved over a 99% accuracy for detecting all types of attacks on the car hacking dataset and a 98% accuracy on the UNSW-NB15 dataset, outperforming other eight intrusion detection techniques.

In cloud computing, service providers rent heterogeneous servers from cloud providers, i.e., Infrastructure as a Service (IaaS), to meet requests of consumers. The heterogeneity of servers and impatience of consumers pose great challenges to service providers for profit maximization. In this article, we transform this problem into a multi-queue model where the optimal expected response time of each queue is theoretically analyzed. A multi-queue request scheduling algorithm framework is proposed to maximize the total profit of service providers, which consists of three components: request stream splitting, requests allocation, and server assignment. A request stream splitting algorithm is designed to split the arriving requests to minimize the response time in the multi-queue system. An allocation algorithm, which adopts a one-step improvement strategy, is developed to further optimize the response time of the requests. Furthermore, an algorithm is developed to determine the appropriate number of required servers of each queue. After statistically calibrating parameters and algorithm components over a comprehensive set of random instances, the proposed algorithms are compared with the state-of-the-art over both simulated and real-world instances. The results indicate that the proposed multi-queue request scheduling algorithm outperforms the other algorithms with acceptable computational time.

Business process monitoring techniques have been investigated in depth over the last decade to enable organizations to deliver process insight. Recently, a new stream of work in predictive business process monitoring leveraged deep learning techniques to unlock the potential business value locked in process execution event logs. These works use Recurrent Neural Networks, such as Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU), and suffer from misinformation and accuracy as they use the last hidden state (as the context vector) for the purpose of predicting the next event. On the other hand, in operational processes, traces may be very long, which makes the above methods inappropriate for analyzing them. In addition, in predicting the next events in a running case, some of the previous events should be given a higher priority. To address these shortcomings, in this paper, we present a novel approach inspired by the notion of attention mechanism, utilized in Natural Language Processing and, particularly, in Neural Machine Translation. Our proposed approach uses all hidden states to accurately predict future behavior and the outcome of individual activities. Experimental evaluation of real-world event logs revealed that the use of attention mechanisms in the proposed approach leads to a more accurate prediction.

Automatic classification algorithms are an important component of expert decision support systems that are used in a number of medical applications including diagnostic radiology and disease detection. This study proposes a deep learning-based framework for medical image classification using wavelet features. Convolutional neural networks are incorporated to discover informative latent patterns and features from a set of X-ray images pertaining to human body parts. The features are then passed to a classifier for labelling the respective X-ray images. The experimental results show that the low-pass filter wavelet-based convolutional model outperforms the original convolutional network and some models for classifying X-ray images. The performance of the proposed method implies that it can be implemented effectively in practice for disease detection using radiological images.

Trust can be employed for finding reliable information in Online Social Networks (OSNs). Since users in OSNs may intentionally change their behavior over time (in some cases for deceiving other users), modeling (pair-wise) trust relations in such complex environment is a challenging task. However, most of the existing trust prediction approaches assume that trust relations are fixed over time and they fail to capture the dynamic behavior of users in OSNs. In this paper, we propose a dynamic deep trust prediction model. As the impact of incidental emotions on trust has been proven in psychology studies, in this paper, we also study this impact on our trust prediction approach. First, we propose a novel deep structure that incorporates users' emotions and their textual contents in OSNs. Second, we use embeddings to represent the users and their self-descriptions provided. Finally, considering different time windows, we dynamically predict pair-wise trust relations. To evaluate our approach, we collected a large twitter dataset. The evaluation results demonstrate the effectiveness of our approach compared to the state-of-the-art approaches.

Business Processes, i.e., a set of coordinated tasks and activities to achieve a business goal, and their continuous improvements are key to the operation of any organization. In banking, business processes are increasingly dynamic as various technologies have made dynamic processes more prevalent. For example, customer segmentation, i.e., the process of grouping related customers based on common activities and behaviors, could be a data-driven and knowledge-intensive process. In this paper, we present an intelligent data-driven pipeline composed of a set of processing elements to move customers' data from one system to another, transforming the data into the contextualized data and knowledge along the way. The goal is to present a novel intelligent customer segmentation process which automates the feature engineering, i.e., the process of using (banking) domain knowledge to extract features from raw data via data mining techniques, in the banking domain. We adopt a typical scenario for analyzing customer transaction records, to highlight how the presented approach can significantly improve the quality of risk-based customer segmentation in the absence of feature engineering.

The rapid and vast adoption of social media and the 21st century technologies is changing the bullying process., i.e., deliberate misuse of power in relationships through social behavior that intends to cause social and psychological harm. In this context, Cyberbullying can be defined as the process of using technology to hurt someone else by sending hurtful messages, pictures, or comments. A successful Cyberbullying detection technique requires an effective combination of big data (generated from social, open, and private data islands), features, and algorithms. While many Cyberbullying detection techniques exist for working with machine learning data and algorithms, support for thinking of new features (to be extracted from images, images' metadata, and textual information generated around images) remains poor. In this paper, we present an intelligent Cyberbullying detection pipeline to: (i) extract features (from an image, image meta-data and textual content generate); (ii) contextualize the extracted features by developing a crowdsourced feedback loop and drawing on human knowledge; and (iii) combining features using a Neural Network to identify and build potentially useful features. We adopt a typical scenario for analyzing social media content generated on Instagram and Twitter to identify Cyberbullying activities. We discuss how the proposed approach significantly improves the quality of extracted knowledge compared to classical natural language analysis techniques.

Enabling the analysis of behavioral disorders over time in social networks, can help in suicide prevention, (school) bullying detection and extremist/criminal activity prediction. In this paper, we present a novel data analytics pipeline to enable the analysis of patterns of behavioral disorders on social networks. We present a Social Behavior Graph (sbGraph) model, to enable the analysis of factors that are driving behavior disorders over time. We use the golden standards in personality, behavior and attitude to build a domain specific Knowledge Base (KB). We use this domain knowledge to design cognitive services to automatically contextualize the raw social data and to prepare them for behavioral analytics. Then we introduce a pattern-based word embedding technique, namely personality2vec, on each feature extracted to build the sbGraph. The goal is to use mathematical embedding from a space with a dimension per feature to a continuous vector space which can be mapped to classes of behavioral disorders (such as cyber-bullying and radicalization) in the domain specific KB. We implement an interactive dashboard to enable social network analysts to analyze and understand the patterns of behavioral disorders over time. We focus on a motivating scenario in Australian government's office of the e-Safety commissioner, where the goal is to empowering all citizens to have safer, more positive experiences online.

Abstract Rule-based systems have been used increasingly to augment learning algorithms for annotating data. Rules alleviate many of the shortcomings inherent in pure algorithmic approaches, in cases algorithms are not working well or lack from enough training data. However, in dynamic curation environments where data are constantly changing, there is a need to craft and adapt rules to keep them applicable and precise. Rule adaptation has been proven to be painstakingly difficult and error-prone, as an analyst is needed for examining the precision of rules and applying different modifications to adapt the imprecise ones. In this paper, we present an autonomic and conceptual approach to adapt data annotation rules. Our approach offloads analysts from adapting rules; it boosts rules to annotate a larger number of items using a set of high-level conceptual features, e.g. topic. We utilize a Bayesian multi-armed-bandit algorithm, an online learning algorithm that adapts rules based on the feedback collects from the curation environment over time. We propose a summarization technique, which offers a set of high-level conceptual features for annotating items by identifying the semantical relationships among them. We conduct experiments on different curation domains and compare the performance of our approach with systems relying on analysts for adapting rules. The experimental results show that our approach has a comparative performance to analysts in adapting rules.

The confluence of technological and societal advances, and more specifically, engagement with Web, social media, and smart devices has the potential to affect the mental behavior of the individuals. Examples include extremist and criminal behaviors such as radicalization and cyber-bullying, which are causing serious issues for humanity. Major barriers to the effective understanding of behavioral disorders on social networks includes the ability to understand the content and context of social documents, as well as the activity of social users. Understanding the patterns of behavioral disorders (e.g., criminal and extremist activities) on social networks, is challenging and requires techniques to contextualize the content of social documents based on the time-aware analysis of personality, behaviour and past activities of social users. In this context, semantic information extraction and enrichment from social documents has the potential to become a vital asset to explore the sign of behavioral disorders and prevent serious issues such as cyber-bullying, suicidal related behavior and radicalization. To address this challenge, in this paper, we present a novel social document analysis pipeline to enable analysts engage with social documents (e.g., a Tweet in Twitter or a post on Facebook) to explore cognitive aspects of behavioral disorders. We implement the pipeline as an extensible and scalable architecture and present the evaluation results.

Customer reviews are now increasingly available on Online Social Networks (OSNs) for a wide range of products and services. Trust in the review's author is a crucial basis for believing in the reliability of reviews generated on such networks. In this context, the main challenge is to predict the unknown trust relationship between two users. Existing trust prediction approaches fail to incorporate textual footprint of users. To address this challenge, we present a deep learning-based graph analytics model to predict trust relations in OSNs. We leverage and extend GraphSAGE, a method for computing node representations in an inductive manner, to develop a deep classifier. We present our experiment with datasets from review websites to train classifiers that predict trust relations between pairs of users, and highlight how our approach significantly improves the quality of predicted trust relations compared to the state-of-the-art approaches.

With Data Science continuing to emerge as a powerful differentiator across industries, organisations are now focused on transforming their data into actionable insights. This task is challenging as in today's knowledge-, service-, and cloud-based economy, businesses accumulate massive amounts of raw data from a variety of sources. Data Lakes introduced as a storage repository to organize this raw data in its native format (supporting from relational to NoSQL DBs) until it is needed. The rationale behind a Data Lake is to store raw data and let the data analyst decide how to cook/curate them later. In this paper, we present the notion of Knowledge Lake, i.e. a contextualized Data Lake. The Knowledge Lake will provide the foundation for big data analytics by automatically curating the raw data in the Data Lake and to prepare them for deriving insights. We present CoreKG-an open source Data and Knowledge Lake service- which offers researchers and developers a single REST API to organize, curate, index and query their data and metadata in the Lake and over time. CoreKG manages multiple database technologies (from Relational to NoSQL) and offers a built-in design for data curation, security and provenance.

Summary In today's knowledge‐, service‐, and cloud‐based economy, an overwhelming amount of business‐related data are being generated at a fast rate daily from a wide range of sources. These data increasingly show all the typical properties of big data: wide physical distribution, diversity of formats, nonstandard data models, and independently managed and heterogeneous semantics. In this context, there is a need for new scalable and process‐aware services for querying, exploration, and analysis of process data in the enterprise because (1) process data analysis services should be capable of processing and querying large amount of data effectively and efficiently and, therefore, have to be able to scale well with the infrastructure's scale and (2) the querying services need to enable users to express their data analysis and querying needs using process‐aware abstractions rather than other lower‐level abstractions. In this paper, we introduce ProcessAtlas, ie, an extensible large‐scale process data querying and analysis platform for analyzing process data in the enterprise. The ProcessAtlas platform offers an extensible architecture by adopting a service‐based model so that new analytical services can be plugged into the platform. In ProcessAtlas, we present a domain‐specific model for representing process knowledge, ie, process‐level entities, abstractions, and the relationships among them modeled as graphs. We provide services for discovering, extracting, and analyzing process data. We provide efficient mapping and execution of process‐level queries into graph‐level queries by using scalable process query services to deal with the process data size growth and with the infrastructure's scale. We have implemented ProcessAtlas as a MapReduce‐based prototype and report on experiments performed on both synthetic and real‐world datasets.

The continuous improvement in connectivity, storage and data processing capabilities allow access to a data deluge from sensors, social-media, news, user-generated, government and private data sources. Accordingly, in a modern data-oriented landscape, with the advent of various data capture and management technologies, organizations are rapidly shifting to datafication of their processes. In such an environment, analysts may need to deal with a collection of datasets, from relational to NoSQL, that holds a vast amount of data gathered from various private/open data islands, i.e. Data Lake. Organizing, indexing and querying the growing volume of internal data and metadata, in a data lake, is challenging and requires various skills and experiences to deal with dozens of new databases and indexing technologies: How to store information items? What technology to use for persisting the data? How to deal with the large volume of streaming data? How to trace and persist information about data? What technology to use for indexing the data? How to query the data lake? To address the above mentioned challenges, we present CoreDB - an open source data lake service - which offers researchers and developers a single REST API to organize, index and query their data and metadata. CoreDB manages multiple database technologies and offers a built-in design for security and tracing.

The Internet of Things (IoT) changes many sectors of our lives. In the healthcare domain, IoT presents as mobile medical applications over various sensors that update healthcare professionals on patients' health information. However, IoT-based healthcare systems also face major challenges in protecting patients' privacy via an effective access control system. This paper presents an ambient home solution framework for privacy-preserving monitoring of patients' health status. We focus on two major points: 1) how to use the data collected from ambient and biometric sensors, to perform the high-level task of activity recognition, and 2) how to secure the collected healthcare data via effective access control. We achieve multi-level access control by using Public Key Infrastructure (PKI) for authentication and Attribute-Based Access Control (ABAC) for authorisation. Our access control system regulates access to healthcare data by classification over healthcare professionals and data. Our system provides guidelines to define data classes and healthcare professional groups and specifies security policies to control access to the data classes. The system is flexible and can incorporate more policy rules, professionals, and data classes.

A large volume of Open Data is being generated on a continuous basis. Examples of this are the case of social, natural, and information systems such as World Wide Web and social networks. Most entities and objects in the Open Data are interconnected, forming a complex, semi-structured, and information-rich networks. In this sense, Linked Open Data has the potential to be similar to a federated database. Since Linked Open Data is based on W3C standards, it is possible to implement a federation infrastructure, however, the current SPARQL standard makes it challenging to analyze the Open Data in an explorative manner. Consequently, it will be hard to discover the hidden knowledge in the relationships among entities in Open Data sources. In this paper, we present Galaxy, a platform for explorative analysis of Open Data Sources. Galaxy facilitates the analysis of Open Data graphs based on simple abstractions, i.e. folders and paths, which enable an analyst to group related entities in the graph or nd paths among entities. Galaxy uses Hadoop data processing platforms to store and retrieve large numbers of RDF triples and to support cost-eective and Web-scale processing of Semantic Web data through a Folder-Path enabled extension of SPARQL.

The Resource Description Framework (RDF) and SPARQL query language are gaining wide popularity and acceptance. In this paper, we present DREAM, a distributed and adaptive RDF system. As opposed to existing RDF systems, DREAM avoids partitioning RDF datasets and partitions only SPARQL queries. By not partitioning datasets, DREAM offers a general paradigm for different types of pattern matching queries, and entirely averts intermediate data shuffling (only auxiliary data are shuffled). Besides, by partitioning queries, DREAM presents an adaptive scheme, which automatically runs queries on various numbers of machines depending on their complexities. Hence, in essence DREAM combines the advantages of the state-of-the-art centralized and distributed RDF systems, whereby data communication is avoided and cluster resources are aggregated. Likewise, it precludes their disadvantages, wherein system resources are limited and communication overhead is typically hindering. DREAM achieves all its goals via employing a novel graph-based, rule-oriented query planner and a new cost model. We implemented DREAM and conducted comprehensive experiments on a private cluster and on the Amazon EC2 platform. Results show that DREAM can significantly outperform three related popular RDF systems.

Worker selection is a significant and challenging issue in crowdsourcing systems. Such selection is usually based on an assessment of the reputation of the individual workers participating in such systems. However, assessing the credibility and adequacy of such calculated reputation is a real challe