In web-based AI services, providers typically host multiple large language models (LLMs) that exhibit diverse capabilities and incur different API costs. Meanwhile, LLM's performance depends not only on its inherent capacity but also on the reasoning strategy it employs, which together influence both answer quality and computational cost. A key challenge is therefore how to adaptively allocate models and strategies to achieve high-quality responses under constrained costs. To address this challenge, we propose Route-To-Reason (RTR), a unified routing framework that simultaneously selects suitable LLMs and reasoning strategies according to query complexity and user budget. Specifically, RTR learns dense vector representations of models and strategies that capture their behavioral characteristics in handling different queries. Leveraging these embeddings, RTR builds a routing table that estimates the cost and performance of different model–strategy pairs. During inference, RTR consults this routing table to dynamically assign the most appropriate pair, enabling adaptive and cost-efficient reasoning tailored to query difficulty and budget scenarios. Extensive experiments across multiple reasoning benchmarks show that RTR achieves comparable or higher accuracy than the best single LLM while substantially reducing both token usage and API cost (by up to 60%), achieving a superior trade-off between performance and efficiency. By lowering the overhead of large-scale LLM inference, RTR contributes to cost-aware and environmentally sustainable deployment of web-based AI services.
论文检索
输入标题、作者或关键词，从 13,513 篇学术成果中精准定位
Federated domain generalization (FDG) for time-series classification (TSC) poses a critical challenge for modern intelligent web services, which rely on edge-collected time-series signals from diverse mobile applications and web devices (e.g., wearables sensors) to support decision-making. The source heterogeneity and temporal dynamics give rise to out-of-distribution (OOD) patterns, which hinder the model's ability to generalize to previously unseen users and devices. In this work, we propose Federated Generalization via Diversity Generation (FedDiG), a diffusion-based FDG framework that captures intra-client distribution shifts from a frequency-domain perspective and employs cross-frequency sampling to synthesize time-series data with diverse spectral patterns. Specifically, FedDiG first performs frequency-proxy representation learning on clients to serve as diffusion conditions. The server then aggregates client-side frequency proxies to construct a global proxy pool and applies class-wise mixup to create novel frequency features. These features guide a global diffusion model to produce diverse data, enabling the simulation of previously unseen patterns and thereby enhancing model training. Extensive experiments on four cross-domain time-series benchmarks demonstrate that FedDiG significantly outperforms state-of-the-art federated learning and FDG baselines, particularly under small-data regimes and large-scale client scenarios, achieving robust generalization to unseen domains in federated settings. This work bridges distribution-diversity synthesis and FDG for time-series to support robust, scalable web applications fed by edge-collected signals, delivering web-scale generalization across heterogeneous web, mobile, and IoT clients.
Web service administrators must ensure the stability of multiple systems by promptly detecting anomalies in Key Performance Indicators (KPIs). Achieving the goal of ''train once, infer across scenarios'' remains a fundamental challenge for time series anomaly detection models. Beyond improving zero-shot generalization, such models must also flexibly handle sequences of varying lengths during inference—ranging from one hour to one week—without retraining. Conventional approaches rely on sliding-window encoding and self-supervised learning, which restrict inference to fixed length inputs. Large Language Models (LLMs) have demonstrated remarkable zero-shot capabilities across general domains. However, when applied to time series data, they face inherent limitations due to context length. To address this issue, we propose ViTs, a Vision-Language Model (VLM)-based framework that converts time series curves into visual representations. By rescaling time series images, temporal dependencies are preserved while maintaining a consistent input size, thereby enabling efficient processing of arbitrarily long sequences without context constraints. Training VLMs for this purpose introduces unique challenges, primarily due to the scarcity of aligned time series image–text data. To overcome this, we employ an evolutionary algorithm to automatically generate thousands of high-quality image–text pairs and design a three-stage training pipeline consisting of: (1) time series knowledge injection, (2) anomaly detection enhancement, and (3) anomaly reasoning refinement. Extensive experiments demonstrate that ViTs substantially enhance the ability of VLMs to understand and detect anomalies in time series data. All datasets and code will be publicly released at: https://github.com/Uni-WangZexin/ViTs.
Cross-view geo-localization (CVGL) establishes correspondences between ground-level and satellite images of the same geographic location, serving as a fundamental technology for smart city applications, including autonomous navigation, urban planning, and location-based services. Current CVGL approaches fall into two categories: feature-based methods achieve superior performance through 2D representation learning but lack interpretability. Spatial-based methods provide geometric understanding and interpretable matching but suffer from limited spatial modeling and weak cross-view alignment, leading to lower performance. We reformulate CVGL from a spatial perspective and propose an auxiliary task-enhanced network. The network captures spatial semantics and provides explicit alignment processes with visualizable results. We introduce an auxiliary spatial semantic alignment (SSA) task that learns spatial structure via vision foundation models (VFM) and BEV transformation to enhance the primary CVGL task. The primary task captures visual semantics, including texture and appearance. Within this unified framework with shared encoders, the primary task enriches the learned embeddings by fusing spatial structure with visual semantics, yielding spatially complete representations. Extensive experiments on three standard CVGL benchmarks demonstrate that our method significantly surpasses previous spatial-based approaches while maintaining competitive performance with state-of-the-art (SOTA) feature-based methods, achieving 98.48% R@1 on CVUSA and 71.05% R@1 on CVACT\_test. We provide comprehensive analyses through pixel-level activation maps and feature-space UMAP visualizations to validate both effectiveness and interpretability.
Generative modeling offers a promising solution to data scarcity and privacy challenges in time series analysis. However, the structural complexity of time series, characterized by multi-scale temporal patterns and heterogeneous components, remains insufficiently addressed. In this work, we propose a structure-disentangled multiscale generation framework for time series. Our approach encodes sequences into discrete tokens at multiple temporal resolutions and performs autoregressive generation in a coarse-to-fine manner, thereby preserving hierarchical dependencies. To tackle structural heterogeneity, we introduce a dual-path VQ-VAE that disentangles trend and seasonal components, enabling the learning of semantically consistent latent representations. Additionally, we present a guidance-based reconstruction strategy, where coarse seasonal signals are utilized as priors to guide the reconstruction of fine-grained seasonal patterns. Experiments on six datasets show that our approach produces higher-quality time series than existing methods. Notably, our model achieves strong performance with a significantly reduced parameter count and exhibits superior capability in generating high-quality long-term sequences. Our implementation is available at https://github.com/decisionintelligence/TimeMAR.
As microservice architecture is evolving toward Logical Data Center (LDC), accurate forecasting of the microservices indicators can support reasonable resource allocation, thereby ensuring the availability and reliability of cloud service. From a macro perspective, due to the architecture hierarchy, microservices exhibit: 1) collaborative relationships derived from shared functionalities, 2) backup relationships between replicas, and 3) dynamic correlation driven by cooperation. From a micro perspective, there exist causal relationships among indicators within a microservice. That is, workload will first impact system consumption, such as CPU and memory usage, then affect service quality like system latency. Based on these insights, we propose MaMiClif, a macro-micro collaborative learning framework for LDC microservice indicators forecasting. MaMiClif constructs Macro Graph and Micro Matrix to model the microservices dependencies and the causality of indicators. To learn fine-grained indicator dependencies, Indicator-Centric Embedding is leveraged to generate representations for indicator series. We use Heterogeneous Graph Convolution to update workload representations based on the Macro Graph, and adopt Causal Sparse Self-attention to integrate causal strength into the self-attention calculation, enabling a comprehensive exploration of dependencies among indicators. Experiments on two datasets, including LDC_MS, which was collected from the LDC system of Ant Group, demonstrate the effectiveness of MaMiClif.
Relational databases (RDBs) underpin the majority of global data management systems, where information is structured into multiple interdependent tables. In social media platforms, for instance, massive user-generated data are organized across related tables such as users, posts, comments, and interactions, enabling large-scale analysis and predictive modeling of social behaviors. To effectively use the knowledge within RDBs for predictive tasks, recent advances leverage graph representation learning to capture complex inter-table relations as multi-hop dependencies. Despite achieving state-of-the-art performance, these methods remain hindered by the prohibitive storage overhead and excessive training time, due to the massive scale of database and the computational burden of intensive message passing across interconnected tables. To alleviate these concerns, we propose and study the problem of Relational Database Distillation (RDD). Specifically, we aim to distill large-scale RDBs into compact heterogeneous graphs while retaining the predictive power (i.e., utility) required for training graph-based models. Multi-modal column information is preserved through node features, and primary–foreign key relations are encoded via heterogeneous edges, thereby maintaining both data fidelity and relational structure. To ensure adaptability across diverse downstream tasks without engaging the traditional, inefficient bi-level distillation framework, we further design a kernel ridge regression-guided objective with pseudo-labels, which produces quality features for the distilled graph. Extensive experiments on multiple real-world RDBs demonstrate that our solution substantially reduces the data size while maintaining competitive performances on classification and regression tasks, creating an effective pathway for scalable learning with RDBs.
They Said Memes Were Harmless - We Found the Ones That Hurt: Decoding Jokes, Symbols, and Cultural References
PDF ↗Meme-based social abuse detection is challenging because harmful intent often relies on implicit cultural symbolism and subtle cross-modal incongruence. Prior approaches, from fusion-based methods to in-context learning with Large Vision-Language Models (LVLMs), have made progress but remain limited by three factors: i) cultural blindness (missing symbolic context), ii) boundary ambiguity (satire vs. abuse confusion), and iii) lack of interpretability (opaque model reasoning). We introduce CROSS-ALIGN+, a three-stage framework that systematically addresses these limitations: (1) Stage I mitigates cultural blindness by enriching multimodal representations with structured knowledge from ConceptNet, Wikidata, and Hatebase; (2) Stage II reduces boundary ambiguity through parameter-efficient LoRA adapters that sharpen decision boundaries; and (3) Stage III enhances interpretability by generating cascaded explanations. Extensive experiments on five benchmarks and eight LVLMs demonstrate that CROSS-ALIGN+ consistently outperforms state-of-the-art methods, achieving up to 17% relative F1 improvement while providing interpretable justifications for each decision.
Social media engagement prediction is a central challenge in computational social science, particularly for understanding how users interact with misinformation. Existing approaches often treat engagement as a homogeneous time-series signal, overlooking the heterogeneous social mechanisms and platform designs that shape how misinformation spreads. In this work, we ask: ''Can neural architectures discover social exchange principles from behavioral data alone?'' We introduce Dreams (Disentangled Representations and Episodic Adaptive Modeling for Social media misinformation engagements), a social exchange theory-guided framework that models misinformation engagement as a dynamic process of social exchange. Rather than treating engagement as a static outcome, Dreams models it as a sequence-to-sequence adaptation problem, where each action reflects an evolving negotiation between user effort and social reward conditioned by platform context. It integrates adaptive mechanisms to learn how emotional and contextual signals propagate through time and across platforms. On a cross-platform dataset spanning 7 platforms and 2.37M posts collected between 2021 and 2025, Dreams achieves state-of-the-art performance in predicting misinformation engagements, reaching a mean absolute percentage error of 19.25%. This is a 43.6% improvement over the strongest baseline. Beyond predictive gains, the model reveals consistent cross-platform patterns that align with social exchange principles, suggesting that integrating behavioral theory can enhance empirical modeling of online misinformation engagement. The source code is available at: https://github.com/ltian678/DREAMS.
The widespread use of social-media graphs has provided a convenient channel for rumor propagation. Rapid localization of rumor sources is therefore crucial for mitigating diffusion and enabling punitive countermeasures. Source Localization (SL) aims to identify the origin nodes given partial infection observations. Although deep-learning-based SL approaches outperform traditional estimators, three fundamental limitations remain: (i) Model Complexity —existing methods enrich node embeddings with cascades of auxiliary features, yielding high-capacity but excessively complex representations, leading to an exponential increase in the number of model parameters; (ii) Annotation gap —to overcome the scarcity of real-world misinformation cascades, current pipelines repeatedly simulate diffusion from a fixed seed, eroding robustness on true, few-shot outbreaks; and (iii) Computational bottleneck —full-model retraining or recurrent cascade simulation is required for every new task, which disqualifies the solutions from real-time deployment. Inspired by the success of prompt learning in NLP and graph learning, we propose LAPS, a Lightweight privilege-Allocation Prompting framework for Source localization. LAPS first trims parameter explosion and data scarcity by pre-training a graph-level source region classifier on adaptive subgraphs with source-prior diffusion data. It then enables few-shot SL via a privilege-allocation prompt module that updates <1% of all the parameters, avoiding model retraining to facilitate efficiency. Extensive experiments on five real-world networks demonstrate the effectiveness and efficiency of our prompt-based framework on few-shot source localization task.
Detecting social bots is critical to ensuring the security of online discourse and maintaining trust in social networks. Early feature-based and text-based methods often fail against bots that mimic human behavior, and graph-based approaches have emerged to better exploit structural signals. However, most existing Graph Neural Networks (GNNs) still focus on pairwise connections, overlooking higher-order relational patterns, and their multi-relation fusion strategies are typically simplistic, ignoring dependencies between relations and user-specific preferences. To overcome these limitations, we propose MPS-Bot, a model that integrates higher-order structure modeling with user-specific cross-relation dependency learning. MPS-Bot introduces a simplex convolutional layer that leverages simplexes derived from network structures to capture group coordination patterns beyond pairwise connections. In addition, a cross-relation dependency attention mechanism adaptively fuses relation-specific representations according to each user's relational preferences, leading to more discriminative and robust multi-relation representations. Extensive experiments on two widely used Twitter bot detection benchmarks, MGTAB and TwiBot-22, show that MPS-Bot generally outperforms state-of-the-art baselines. These findings highlight the effectiveness of higher-dimensional message passing over simplexes as a powerful approach to unmasking bots in social networks.
Dynamic link prediction plays a crucial role in diverse applications including social network analysis, communication forecasting, and financial modeling. While recent Transformer-based approaches have demonstrated promising results in temporal graph learning, their performance remains limited when capturing complex multi-scale temporal dynamics. In this paper, we propose TFWaveFormer, a novel Transformer architecture that integrates temporal-frequency analysis with multi-resolution wavelet decomposition to enhance dynamic link prediction. Our framework comprises three key components: (i) a temporal-frequency coordination mechanism that jointly models temporal and spectral representations, (ii) a learnable multi-resolution wavelet decomposition module that adaptively extracts multi-scale temporal patterns through parallel convolutions, replacing traditional iterative wavelet transforms, and (iii) a hybrid Transformer module that effectively fuses local wavelet features with global temporal dependencies. Extensive experiments on benchmark datasets demonstrate that TFWaveFormer achieves state-of-the-art performance, outperforming existing Transformer-based and hybrid models by significant margins across multiple metrics. The superior performance of TFWaveFormer validates the effectiveness of combining temporal-frequency analysis with wavelet decomposition in capturing complex temporal dynamics for dynamic link prediction tasks. The code is available at https://github.com/SEUFHTong/TFWaveFormer.
Illicit activities such as financial fraud and fake promotions are increasingly prevalent on social applications. The diverse behavioral structures of different types of illicit accounts pose difficulties in designing unified detection strategies. Unlike existing studies that focus on binary detection and often rely on user static profiles, we introduce a novel task: illicit account detection based solely on user behavior sequences. This task presents two key challenges: 1) illicit accounts often mimic benign users by performing normal-looking behavior subsequences, and 2) behaviors with the same action (e.g., add-friend, initiate-transaction) can serve different purposes, such as illicit or benign. To address these challenges, we propose a Pattern-aware Illicit Accounts Detection (PIAD) framework that consists of three components: 1) a dual-perspective pattern mining module that extracts category-specific self- and interaction-behavior patterns from behavior sequences to capture distinct behavioral regularities across different user types; 2) a contextualized action semantic encoding algorithm that aligns action codings with contextual dependencies among behaviors within user sequences to capture variations in purposes when behaviors with the same actions occur under different contexts; and 3) a pattern-aware fusion model that integrates the mined patterns with the context and interaction in behavior sequences to learn discriminative representations for detection. Extensive experiments on real-world datasets demonstrate that PIAD consistently outperforms state-of-the-art baselines with an average 7.51% improvement on F1 score.
Multimodal desire understanding, a task closely related to both emotion and sentiment that aims to infer human intentions from visual and textual cues, is an emerging yet underexplored task in affective computing with applications in social media analysis. Existing methods for related tasks predominantly focus on mining verbal cues, often overlooking the effective utilization of non-verbal cues embedded in images. To bridge this gap, we propose a Symmetrical Bidirectional Multimodal Learning Framework for Desire, Emotion, and Sentiment Recognition (SyDES). The core of SyDES is to achieve bidirectional fine-grained modal alignment between text and image modalities. Specifically, we introduce a mixed-scaled image strategy that combines global context from low-resolution images with fine-grained local features via masked image modeling (MIM) on high-resolution sub-images, effectively capturing intention-related visual representations. Then, we devise symmetrical cross-modal decoders, including a text-guided image decoder and an image-guided text decoder, which enable mutual reconstruction and refinement between modalities, facilitating deep cross-modal interaction. Furthermore, a set of dedicated loss functions is designed to harmonize potential conflicts between the MIM and modal alignment objectives during optimization. Extensive evaluations on the MSED benchmark demonstrate the superiority of our approach, which establishes a new state-of-the-art performance with 1.1% F1-score improvement in desire understanding. Consistent gains in emotion and sentiment recognition further validate its generalization ability and the necessity of utilizing non-verbal cues. Our code is available at: https://github.com/especiallyW/SyDES.
Graph neural networks (GNNs) have become the cornerstone of graph representation learning in real-world applications, especially in social networks such as Facebook and Twitter, where user interactions naturally form graph-structured data. To safeguard sensitive information, including user attributes and personal connections, locally private graph learning protocols employ local differential privacy (LDP) to provide rigorous user-level guarantees. Despite these protections, we show that such protocols remain highly vulnerable to data poisoning attacks. Our approach targets the full-privacy setting, where both node features and edges are LDP-protected, and executes coordinated manipulations that significantly degrade utility, such as node classification accuracy, across multiple social network benchmarks. To counter these threats, we propose CureNet, a defense framework with four key components: (1) local data perturbation for privacy, (2) trimmed screening to filter abnormal submissions, (3) privacy-aware fake node pruning to remove sophisticated adversaries, and (4) a utility enhancement module to recover graph learning performance under privacy constraints. Extensive experiments on four real-world social network datasets—LastFM, Twitch, Github, and Facebook—demonstrate both the severe impact of our attack and the effectiveness of our defense. This study underscores the critical need to secure privacy-preserving graph learning against sophisticated poisoning threats.
Spamming activities (e.g., fake reviews, click farming, and deceptive content promotion) are increasingly conducted through collusive groups that exploit collective dynamics to manipulate platform metrics and mislead users, posing serious threats to the fairness, credibility, and functionality of online systems. To counteract these harmful behaviors, the task of spam detection has emerged as a critical area of research. However, existing detection methods generally remain limited in three key aspects: (i) They treat detection as a standard classification task, where representation learning and optimization are loosely coupled and suboptimal for capturing complex behaviors; (ii) They rely primarily on individual-level representation modeling, making it difficult to detect collective cheating strategies; (iii) They lack dedicated objective functions explicitly designed to characterize group-level spamming activities. To overcome these limitations, we introduce a collusion-aware Set-level learning framework (SetDet) that redefines the spam Detection task as a unified setwise optimization problem. Our approach offers three core advantages: (i) It enables end-to-end optimization by jointly learning representations and performing detection in a single, integrated process; (ii) It incorporates a model-level design for collusion representation, effectively capturing the temporal and relational patterns of coordinated spam; (iii) It pioneers a dedicated set-level optimization criterion that aligns closely with the structural characteristics of group-based cheating behaviors and accounts for class imbalance in real-world scenarios. Extensive experiments confirm the generalizability and superior performance of our framework across diverse spam scenarios and collusion strategies.
Information diffusion prediction aims to forecast the temporal spread of opinions and behaviors by identifying potential adopters. Existing methods typically treat information diffusion as a sequence of individual adoptions and rely on computationally expensive pairwise (one-to-one) influence computations, often restricting predictions to just the next adopter. This individual-level paradigm both misrepresents real-world collective (many-to-many) influences and suffers a critical efficiency trade-off: to remain feasible, such models must truncate long diffusion histories, thereby overlooking early initiators and opinion leaders. To overcome these limitations, we formalize a more practical task: Group-based Information Diffusion Prediction, and propose an effective and scalable GRID framework. Specifically, GRID first learns group-oriented graph embeddings via a task-regularized information bottleneck objective, which amplifies key influence pathways and produces reliable user embeddings for group identification. Built on these embeddings, the core GroupAttn module captures inter-group influence while reducing complexity from quadratic to linear in cascade length. This enables the modeling of ultra-long cascades (exceeding 10,000 users) without truncation while preserving representational fidelity within a provable error bound. Finally, a group-wise objective guides the model to predict semantically meaningful future groups. Extensive experiments on four real-world datasets show that GRID outperforms ten state-of-the-art baselines by an average of 10.65% in accuracy, while achieving an order-of-magnitude gain in efficiency and extending the supported cascade length by up to 10 times.
Taxonomy completion is the task of integrating new concepts into an existing taxonomy by determining the appropriate hypernym--hyponym relations. Existing approaches often struggle with the inherent imbalance between leaf and non-leaf edges, which induces bias in representation learning. In this paper, we propose BLEND: Balanced and Leaf-Enhanced Dual Fine-Tuning for Taxonomy Completion, a novel framework designed to mitigate this inductive bias. Our method employs independent fine-tuning of two lightweight large language models (LLMs): one optimized with a leaf-focused objective and the other trained with a balanced focused strategy. To further enhance structural understanding, we apply contrastive learning over structure-encoded paths and introduce a combined loss function, enabling more robust representation of hierarchical relations. Extensive experiments on three real-world benchmark datasets demonstrate that BLEND achieves up to 9.32% improvement in recall or hit metrics compared to state-of-the-art approaches. Moreover, BLEND delivers efficient inference while outperforming the latest baseline COMI, highlighting its effectiveness for taxonomy completion tasks.
Recent advances in knowledge representation learning (KRL) highlight the urgent necessity to unify symbolic knowledge graphs (KGs) with language models (LMs) for richer semantic understanding. However, existing approaches typically prioritize either graph structure or textual semantics, which leaves a gap, i.e., a unified framework that simultaneously captures global KG connectivity, nuanced linguistic context, and discriminative reasoning semantics. To bridge this gap, we introduce KG-BiLM, a bidirectional LM framework that fuses structural cues from KGs with the semantic expressiveness of generative transformers. KG-BiLM incorporates three key components: (i) Bidirectional Knowledge Attention, which eliminates the causal mask to enable full interaction among all tokens and entities; (ii) Knowledge-Masked Prediction, which encourages the model to leverage both local semantic contexts and global graph connectivity; and (iii) Contrastive Graph Semantic Aggregation, which preserves KG structure via contrastive alignment of sampled subgraph representations. Extensive experiments on standard benchmarks demonstrate that KG-BiLM outperforms the state-of-the-art baselines in link prediction, especially on large-scale graphs with complex multi-hop relations—validating its effectiveness in unifying structural information and textual semantics. The source code of KG-BiLM is available at https://github.com/zirui-chen/kg-bilm
Accurate representation of multimodal knowledge is crucial for event forecasting in real-world scenarios. However, existing studies have largely focused on static settings, overlooking the dynamic acquisition and fusion of multimodal knowledge. 1) At the knowledge acquisition level, how to learn time-sensitive information of different modalities, especially the dynamic structural modality. Existing dynamic learning methods are often limited to shallow structures across heterogeneous spaces or simple unispaces, making it difficult to capture deep relation-aware geometric features. 2) At the knowledge fusion level, how to learn evolving multimodal fusion features. Existing knowledge fusion methods based on static coattention struggle to capture the varying historical contributions of different modalities to future events. To this end, we propose DyMRL, a Dynamic Multispace Representation Learning approach to efficiently acquire and fuse multimodal temporal knowledge. 1) For the former issue, DyMRL integrates time-specific structural features from Euclidean, hyperbolic, and complex spaces into a relational message-passing framework to learn deep representations, reflecting human intelligences in associative thinking, high-order abstracting, and logical reasoning. Pretrained models endow DyMRL with time-sensitive visual and linguistic intelligences. 2) For the latter concern, DyMRL incorporates advanced dual fusion-evolution attention mechanisms that assign dynamic learning emphases equally to different modalities at different timestamps in a symmetric manner. To evaluate DyMRL's event forecasting performance through leveraging its learned multimodal temporal knowledge in history, we construct four multimodal temporal knowledge graph benchmarks. Extensive experiments demonstrate that DyMRL outperforms state-of-the-art dynamic unimodal and static multimodal baseline methods.