The rapid growth of online short texts has made specialized analysis essential, as these texts are sparse and information-limited. Short text clustering (STC) is critical for automatically grouping unlabeled texts into meaningful clusters, supporting applications such as sentiment analysis, spam filtering, and social media personalization. In the context of massive online content, deep clustering seeks to uncover semantic categories by measuring distances in the representation space. Consequently, aligning clustering pseudo-labels with the true category distribution is crucial for effective self-supervised training, particularly under class imbalance and distribution skew commonly observed in web data. To address this challenge, we propose the Cluster Identification-Guided Dual Correction (CIDC) framework, which generates reliable pseudo-labels to guide deep clustering. Specifically, given cluster partitions and model-estimated class distributions, we perform Cluster Category Identification (CCI) at each training epoch to determine the most probable category for each cluster. This identification provides the foundation for the Pseudo-Label Correction (PLC) and Prototype-Based Correction (PBC) modules, which jointly enhance pseudo-label reliability and representation learning. In the PLC module, samples whose model-estimated class distributions conflict with the assigned cluster category are corrected, thereby improving semantic alignment within clusters. In the PBC module, representative and reliable prototypes are selected according to cluster categories and model predictions to guide training, further strengthening representation discriminability. Extensive experiments demonstrate that CIDC consistently outperforms existing methods in terms of clustering accuracy and mutual information, particularly in unsupervised settings characterized by class imbalance and noisy data.
论文检索
输入标题、作者或关键词，从 100,903 篇学术成果中精准定位
Multimodal Sentiment Analysis (MSA) and Emotion Recognition in Conversations (ERC) have rapidly developed into pivotal tasks in artificial intelligence. Large Language Models (LLMs) offer powerful semantic reasoning and computational capabilities, showing great potential for understanding emotional content. However, when applied to multimodal sentiment data, LLMs face significant challenges, including the inability to directly process heterogeneous data, difficulties in coping with feature misalignment and suboptimal cross-modal fusion. To address these challenges, we propose a novel multimodal sentiment inference framework named ATGFB-MFF which grounded in fiber bundle theory. This method decomposes multimodal features into an adaptive text-guided shared semantic space and fiber offset spaces to achieve structured alignment and fusion. Then the fused features are converted into structured pseudo-token sequences for effective inference via frozen LLMs. We also introduce two loss functions respectively called shared space consistency loss and fiber offset regularization loss which are used to improve representation stability. Extensive experiments on four benchmark datasets demonstrate that ATGFB-MFF consistently outperforms state-of-the-art baselines. These results highlight the efficacy of geometric structural modeling in unlocking the potential of LLMs for multimodal sentiment inference.
Multi-view multi-label classification (MvMLC) is indispensable for modern web applications aggregating information from diverse sources. However, real-world web-scale settings are rife with missing views and continuously emerging classes, which pose significant obstacles to robust learning. Prevailing methods are ill-equipped for this reality, as they either lack adaptability to new classes or incur exponential parameter growth when handling all possible missing-view patterns, severely limiting their scalability in web environments. To systematically address this gap, we formally introduce a novel task, termed incomplete multi-view multi-label class incremental learning (IMvMLCIL), which requires models to simultaneously address heterogeneous missing views and dynamic class expansion. To tackle this task, we propose E2PL, an Effective and Efficient Prompt Learning framework for IMvMLCIL. E2PL unifies two novel prompt designs: task-tailored prompts for class-incremental adaptation and missing-aware prompts for the flexible integration of arbitrary view-missing scenarios. To fundamentally address the exponential parameter explosion inherent in missing-aware prompts, we devise an efficient prototype tensorization module, which leverages atomic tensor decomposition to elegantly reduce the prompt parameter complexity from exponential to linear w.r.t. the number of views. We further incorporate a dynamic contrastive learning strategy explicitly model the complex dependencies among diverse missing-view patterns, thus enhancing the model's robustness. Extensive experiments on three benchmarks demonstrate that E2PL consistently outperforms state-of-the-art methods in both effectiveness and efficiency. The codes and datasets are available at https://anonymous.4open.science/r/code-for-E2PL.
Multimodal sentiment analysis (MSA) has witnessed remarkable advancements in recent years. Existing MSA methods focus primarily on learning coarse-grained representations from different modalities to perform global cross-modal alignment or fusion. However, these approaches often neglect fine-grained valuable sentimental clues derived from local cross-modal interactions. Furthermore, the cross-modal alignment and fusion of complex global and local cross-modal information pose significant challenges in MSA tasks. To address this issue, we propose a novel MSA framework that simultaneously captures coarse-grained and fine-grained cross-modal sentiment cues through global and local cross-modal alignment and fusion. Our approach consists of three key components: i) optimal transport-based global and local cross-modal alignment, which separately aligns valuable global and local sentiment clues across modalities, ii) global and local cross-modal gated attention, which respectively fuse the aligned global and local cross-modal representations, and iii) prototype-informed information bottleneck, which utilizes learnable sentiment prototypes and contrastive prototype match to eliminate redundant cross-modal information at both global and local levels. Extensive experiments conducted on two publicly available MSA datasets demonstrate the effectiveness and superiority of our proposed model.
Although Large Language Models (LLMs) have demonstrated significant proficiency in code generation, their monolithic and correlation-driven nature renders them susceptible to systematic cognitive biases, deficient counterfactual reasoning, and adversarial manipulation—a characteristic we term cognitive vulnerability. Such vulnerabilities compromise the reliability and security of AI-assisted software development, potentially resulting in code that is not only functionally incorrect but also biased, insecure, and difficult to validate using conventional testing paradigms. While recent multi-agent systems enhance workflow efficiency via task decomposition, they do not fundamentally address these reasoning deficits. This highlights the need for a framework capable of proactively identifying and mitigating the cognitive flaws of an LLM during the reasoning process. To address this challenge, we introduce CodeForge, a multi-agent adversarial reasoning framework that reframes code generation as a cognitive crucible. This process involves a structured debate among three specialized agents—an Optimist, a Pragmatist, and an Adversarial Skeptic—who iteratively cross-examine and refine solution plans. Following convergence, an adversarial verification module systematically generates counterfactual perturbations to stress-test and enhance the final plan. Comprehensive evaluations on the HumanEval and MBPP benchmarks demonstrate that CodeForge significantly outperforms state-of-the-art methods, achieving a pass@1 of 97.3% with GPT-4. Ablation studies confirm the necessity of both the adversarial dialogue and counterfactual verification components. This work represents a shift from passive debugging to proactive cognitive hardening, establishing a pathway toward more trustworthy automated software engineering.
While Large Language Models (LLMs) excel at many reasoning tasks, their native inability to produce calibrated, multi-class probability distributions limits their use in high-stakes Web applications like content moderation and fraud detection. Existing methods to elicit probabilities from LLMs either sacrifice their crucial Chain-of-Thought (CoT) reasoning capabilities or suffer from poor calibration. To address this, we introduce a new paradigm, Verbalized Probability Distribution, and a novel training framework, RLVP (Reinforcement Learning with Verbalized Probabilities). RLVP fine-tunes an LLM to generate both an interpretable CoT and a complete, verbalized probability distribution. We overcome the ''insufficient reward granularity'' problem in standard Reinforcement Learning (RL) for classification by using soft probabilities from expert tabular models as a dense reward curriculum. Through large-scale joint training on 169 tabular tasks, we demonstrate that a single RLVP-trained model can surpass a strong, task-specific XGBoost baseline on up to 55% of tasks. More importantly, the trained model achieves state-of-the-art few-shot performance on unseen, heterogeneous Web benchmarks that mix structured data with free text, achieving performance comparable to or superior than expert models trained on the same limited data. This showcases a strong capability for generalization and knowledge transfer to complex Web data. Our work presents a viable path toward building general-purpose, probabilistically-sound, and interpretable foundation models for the Web.
Generative latent diffusion models (LDMs) have been extensively applied in various fields yet underperform in time-series prediction. Therefore, We propose the Re-Diffusion model, a latent diffusion approach that generates backbone residuals specifically tailored for time-series forecasting. The model comprises a variational autoencoder that compresses the residuals between the actual future values and the predictions from the backbone into latent space. It also includes a conditional diffusion generator to forecast the potential distribution of these residuals. Our findings reveal that this latent-space methodology particularly enhances existing backbone predictors, by effectively reducing prediction bias through an advanced estimation of complex error distributions. While previous diffusion-based models tend to struggle with long-term forecasting, Re-Diffusion integrates the strengths of diffusion methods, leading to improvements in long-term predictions. Our experimental results indicate that the Re-Diffusion model achieves a 10% promotion over state-of-art predictors, marking a significant advancement in the field of time-series forecasting.
Knowledge Tracing (KT) is a core task in intelligent tutoring systems, designed to model the dynamic evolution of students' knowledge states by predicting their performance on specific problems. However, existing KT models, primarily developed for traditional subject domains, exhibit poor generalization to programming tasks due to several distinct challenges. First, student responses in programming tasks are characterized by high uncertainty. Second, programming tasks typically provide continuous scores based on the proportion of passed test cases, in contrast to the binary correctness assumed by traditional KT models, with these test-case-based scores often being noisy. Third, most KT models rely exclusively on the understanding of knowledge concepts for performance prediction, whereas success in programming tasks is heavily contingent upon logical reasoning and problem-specific comprehension. To address these issues, we propose an LLM-driven Interaction Enrichment Framework (MIE) to mitigate high uncertainty and problematic labeling, and introduce the Multi-Level Programming Knowledge Tracing (MLPKT) model to capture students' knowledge states across multiple dimensions. MLPKT conducts multi-layer analysis of student submissions to identify the root causes of errors and assign semantically meaningful fine-grained labels. Additionally, we propose a three-level, three-phase KT architecture that captures knowledge dynamics across three dimensions—problems, concepts, and logical skills—through the phases of learning, forgetting/reinforcement, and application. Extensive experiments on three datasets demonstrate that MIE+MLPKT consistently outperforms 18 baseline methods. Our code is available at: https://anonymous.4open.science/r/MIE-MLPKT-D654.
As an important task in modern web technologies, multivariate time series forecasting drives many core functionalities. However, in realistic web-scale environments, sensor failures, privacy filtering, sampling, and instrumentation churn frequently yield missing variables, making training sets complete while test sets contain only a small subset of variables. The challenge lies in utilizing incomplete data for forecasting, which is known as Variable Subset Forecasting (VSF). Distribution shift is inherent to time series and remains in VSF, including inter-series shift as changes in cross-series correlations and intra-series shift as substantial distribution differences within the same series across different time windows. Existing VSF approaches typically impute missing variables and then forecast on the completed series, yet they overlook these shifts and thus underperform in realistic web-scale scenarios. To address these challenges, we propose Shift Resilient Diffusive Imputation (SRDI), a framework tailored to VSF and robust to distribution shift. Specifically, SRDI integrates a divide-conquer strategy with the denoising process, which decomposes the input into invariant patterns and variant patterns, representing the temporally stable parts of inter-series correlation and the highly fluctuating parts, respectively. By extracting spatiotemporal features from each part separately and then appropriately combining them, inter-series shift can be effectively mitigated. Then, we innovatively organize SRDI and the forecasting model into a meta-learning paradigm tailored for VSF scenarios. We address the intra-series shift by treating time windows as tasks during training and employing an adaptation process before testing, which naturally supports robust online forecasting in dynamic web environments. Extensive experiments on four datasets have demonstrated our superior performance compared with state-of-the-art methods. Our code is available at https://github.com/xhhmacau/SRDI.
The strategic planning and reliability of modern web services, from cloud infrastructures to e-commerce platforms, increasingly hinge on accurate long-term forecasting of high-dimensional time series. A fundamental challenge within this task is modeling the intricate point-to-point dependencies that span across both time and variable dimensions. However, many existing methods face restricted direction modeling and computational inefficiency due to their reliance on localized paradigms and Transformer architectures. To address these, we propose replacing Self-Attention with autocorrelation, achieving two key innovations: 1) We propose calculating autocorrelation across both variable and time dimensions, which is a global paradigm, to model point-to-point dependencies. 2) Our proposed Spectral Product Mechanism (SPM) optimizes traditional autocorrelation into a data-driven form suitable for deep learning. Moreover, SPM reformulates autocorrelation as spectral product and reduces the complexity from O(N2) to O(NlogN), while its Hadamard product-based correlation score matrix further reduces core computation to O(N) compared to Self-Attention's O(N2) matrix multiplication. We further propose a Generalized Spectral Product Mechanism (GSPM), which extends traditional autocorrelation by mapping input into distinct feature representations, enabling modeling of complex dependencies through cross-feature correlations. SPM and GSPM surpass current state-of-the-art (SOTA) methods on 14 authoritative benchmarks, collectively securing the top rank on 22 out of 28 metrics, while ranking 1st in time, 2nd in memory, and 2nd in parameter overhead. Source code is available at: https://github.com/lxy-PhD2022/SPM.
Low-rank adaptation (LoRA) has emerged as an efficient fine-tuning technique for large language models, enabling parameter-efficient updates while maintaining task performance. However, LoRA suffers from two key issues: 1) inefficient feature learning when the width n (embedding dimension) is large, and 2) ineffective updates to the adapter matrix A due to the initialization of B as zero. We propose LoRA-E2, which utilizes a Gaussian initialization with variance Θ(n-3/4 ) for A, and employs the Gauss-Seidel iteration to train B and A. We theoretically show that LoRA-E2 enables more stable and efficient feature learning with effective parameter updates over standard LoRA. Empirically, LoRA-E2 achieves consistent gains in both natural language understanding and generation tasks. On the GLUE benchmark with T5-base, it improves performance by 1–10% over LoRA. When fine-tuning LLaMA 2-7B on MetaMathQA with GSM8K as validation, LoRA-E2 surpasses LoRA by 1–2% and converges up to ∼3× faster. Code is available at https://github.com/whu-totemdb/LoRA-E2.
Web-scale content is rich in modalities yet frequently incomplete due to device limits, transmission errors, or privacy controls, making learning with missing modalities a core challenge. Prior reconstruction and alignment strategies often fail to preserve a stable class geometry when inputs are partial, leading to anchor drift -- a shift of class prototypes between complete and incomplete views that distorts the shared representation space and degrades generalization. We introduce HiCoD (Hierarchical Consistency-Guided Pro mpt Distillation), which learns a robust, class-anchored semantic space. HiCoD combines: (1) a modality-aware semantic graph that restores cross-modal structure under partial observations; (2) dual-level anchoring that unifies large-language-model–derived global category prototypes with top-K local exemplars to balance cross-modal coherence and modality-specific detail; and (3) multi-level distillation that aligns unimodal features, fused embeddings, and prompt-completed signals within a single anchor space. Across CMU-MOSI, CMU-MOSEI, and additional benchmarks, HiCoD sets a new state of the art under both fixed-pattern and random missingness, improving Acc-2 by up to 6.4 points over MPLMM and remaining robust when key modalities are absent.
The prevalence of sarcasm in multimodal dialogues on the social platforms presents a crucial yet challenging task for understanding the true intent behind online content. Comprehensive sarcasm analysis requires two key aspects: Multimodal Sarcasm Detection (MSD) and Multimodal Sarcasm Explanation (MuSE). Intuitively, the act of detection is the result of the reasoning process that explains the sarcasm. Current research predominantly focuses on addressing either MSD or MuSE as a single task. Even though some recent work has attempted to integrate these tasks, their inherent causal dependency is often overlooked. To bridge this gap, we propose MuVaC, a variational causal inference framework that mimics human cognitive mechanisms for understanding sarcasm, enabling robust multimodal feature learning to jointly optimize MSD and MuSE. Specifically, we first model MSD and MuSE from the perspective of structural causal models, establishing variational causal pathways to define the objectives for joint optimization. Next, we design an alignment-then-fusion approach to integrate multimodal features, providing robust fusion representations for sarcasm detection and explanation generation. Finally, we enhance the reasoning trustworthiness by ensuring consistency between detection results and explanations. Experimental results demonstrate the superiority of MuVaC in public datasets, offering a new perspective for understanding multimodal sarcasm.
Traffic prediction plays a pivotal role in contemporary web technologies, motivating various intelligent web services such as route planning and remote traffic management. Many recent proposals that target deep learning for traffic prediction solely leverage historical traffic observations to predict future ones. However, traffic prediction is always susceptible to different factors such as road networks and social events, exhibiting different modalities. Most existing methods focus on a single modality, failing to capture the comprehensive traffic patterns among various factors, resulting in sub-optimal performance. Web-sourced geo-images, e.g., satellite imagery, encompass comprehensive contextual information and offer an effective way to represent diverse modalities. To unleash the power of such geo-images, we propose VisionST, a Vision-augmented Spatial-Temporal Neural Network, which coordinates cross-modal traffic prediction with interactive geo-image encoding. To bolster resilience against highly intricate and overlapping traffic patterns, VisionST features a visual semantic extraction mechanism and a pattern-guided aggregation mechanism. The former extracts node-level visual tokens and node-to-node visual relation patterns from geo-referenced images. The latter generates relation patterns that encompass visual, spatial, and temporal aspects, constraining nodes to interact with these relation patterns for contextual information interaction. Extensive experiments on real large-scale datasets offer insight into the effectiveness of the proposed solutions, showing that VisionST consistently outperforms state-of-the-art baselines.
Time series anomaly detection aims to identify samples that deviate from a normal sample distribution in a time series, enabling various web-centric applications. Most existing approaches are static, targeting pre-defined types of anomalies. These methods thus fail to work well on streaming time series with changing data distributions and anomaly formats. To contend with such streaming time series and to accommodate memory constraints, we propose the first data-efficient streaming time series anomaly detection framework, called DESS. To accumulate historical knowledge, DESS includes a novel evolving proxy generation module to synthesize a small but informative proxy summarizing the historical data, facilitating data efficiency. Next, DESS employs an innovative heterogeneous temporal feature extraction module to explicitly capture correlations of multi-level time series semantics. Finally, DESS enables fast streaming anomaly detection by employing a parameter-efficient training scheme that only activates a subset of lightweight parameters while ensuring performance. Extensive experiments on real data offer insight into the effectiveness and efficiency of DESS, showing that it is able to outperform the best baselines by up to 17.53% while reducing the training time by up to 64.88%.
Towards Multi-Label Text Interpretation with Chain-of-Thought Prompting and Contextualized Knowledge
Existing multi-label topic models face several challenges when interpreting texts annotated with multiple labels: (1) they often associate irrelevant text segments with incorrect labels, which negatively impacts both segment and label interpretation; (2) they fail to effectively capture the semantic relationships between tokens and labels within the segment; (3) they do not integrate contextualized knowledge that could improve interpretability. To overcome these issues, we introduce the Contextualized Prompting Topic Model (CPTM). CPTM utilizes Chain-of-Thought (CoT) prompting to better align text segments with their semantically relevant labels. Furthermore, it integrates label-specific token visualization and topic mining procedure to facilitate the interpretation of tokens and labels. Experimental evaluations conducted on three multi-label text datasets show that CPTM significantly outperforms existing models in both segment and label interpretation. Human assessments also verify CPTM's effectiveness in accurately identifying label-relevant tokens within segments and providing insightful token-level interpretation.
Transformer-based foundation models have achieved remarkable progress in tasks such as time-series forecasting and image segmentation. However, they frequently suffer from error accumulation in multivariate long-sequence prediction and exhibit vulnerability to out-of-distribution samples in image-related tasks. Furthermore, these challenges become particularly pronounced in large-scale Web data analysis tasks, which typically involve complex temporal patterns and multimodal features. This complexity substantially increases optimization difficulty, rendering models prone to stagnation at saddle points within high-dimensional parameter spaces. To address these issues, we propose a lightweight Transformer architecture in conjunction with a novel Escape-Explore Optimizer (EEO). The optimizer enhances both exploration and generalization while effectively avoiding sharp minima and saddle-point traps. Experimental results show that, in representative Web data scenarios, our method achieves performance on par with state-of-the-art models across 11 time-series benchmark datasets and the Synapse medical image segmentation task. Moreover, it demonstrates superior generalization and stability, thereby validating its potential as a versatile cross-task foundation model for Web-scale data mining and analysis.
Recent advances in generative artificial intelligence, particularly large language models (LLMs), have opened new opportunities for enhancing recommender systems (RecSys). Most existing LLM-based RecSys approaches operate in a discrete space, using vector-quantized tokenizers to align with the inherent discrete nature of language models. However, these quantization methods often result in lossy tokenization and suboptimal learning, primarily due to inaccurate gradient propagation caused by the non-differentiable argmin operation in standard vector quantization. Inspired by the emerging trend of embracing continuous tokens in language models, we propose ContRec, a novel framework that seamlessly integrates continuous tokens into LLM-based RecSys. Specifically, ContRec consists of two key modules: a σ-VAE Tokenizer, which encodes users/items with continuous tokens; and a Dispersive Diffusion module, which captures implicit user preference. The tokenizer is trained with a continuous Variational Auto-Encoder (VAE) objective, where three effective techniques are adopted to avoid representation collapse. By conditioning on the previously generated tokens of the LLM backbone during user modeling, the Dispersive Diffusion module performs a conditional diffusion process with a novel Dispersive Loss, enabling high-quality user preference generation through next-token diffusion. Finally, ContRec leverages both the textual reasoning output from the LLM and the latent representations produced by the diffusion model for Top-K item retrieval, thereby delivering comprehensive recommendation results. Extensive experiments on four datasets demonstrate that ContRec consistently outperforms both traditional and state-of-the-art LLM-based recommender systems. Our results highlight the potential of continuous tokenization and generative modeling for advancing the next generation of recommender systems.
Dynamic Facial Expression Recognition (DFER), as a crucial part of affective computing, has broad applications in many areas such as human-computer interaction and social media content analysis. Effectively integrating multimodal information, particularly audio-visual signals, remains the core challenge. However, existing approaches are generally constrained by two major challenges: (1) shallow and static fusion mechanisms, which fail to capture the dynamic co-evolution of audio-visual features during deep interaction; (2) implicit and coarse alignment strategies, which are insufficient to bridge the modality gap caused by heterogeneous feature distributions. To address these issues, we propose a novel framework, BHGap, which integrates deep iterative prompt generation with multi-stage feature alignment and fusion. The key idea is to reformulate audio-visual collaboration from a one-shot fusion event into a continuous, reciprocal generation process that spans every layer of frozen backbone encoders. Specifically, we design a State Space Model (SSM)-based cross-modal prompt generator that dynamically produces ''guidance prompts'' for the counterpart modality at each encoding layer, thereby enabling deep and fine-grained feature co-evolution. Beyond encoding, we further introduce a coarse-to-fine multi-stage alignment module: at the macro level, low-rank adversarial alignment is employed to establish spatio-temporal congruity between audio and video while reducing global distributional discrepancies; at the micro level, Maximum Mean Discrepancy (MMD) constraints combined with implicit differentiation optimization ensure fine-grained statistical consistency and semantic alignment. Extensive experiments on the public DFEW and MAFW datasets demonstrate that our method achieves state-of-the-art performance, offering a new paradigm of deep iterative fusion and explicit alignment for multimodal emotion recognition. Code is available at https://github.com/NDYZD666/-public-BHGap.
Insider threat detection (ITD) is notoriously difficult: malicious actions are rare, context-dependent, and deliberately hidden within massive volumes of legitimate user behavior. Existing ITD methods rely on single- or fused-view models, which lack extensibility and therefore fail to leverage the supervisory signals from newly introduced complementary views. While ensembling is a natural next step, its direct application to ITD confronts three core obstacles: scalability bottlenecks from independently trained sub - models, semantic misalignment across heterogeneous feature spaces, and view imbalance, where strong views overshadow weaker yet informative ones. In this work, we propose Insight-LLM, the first extensible multi-view fusion framework tailored for ITD. Insight-LLM encodes each view with frozen pre-trained backbones and aligns heterogeneous representations into a unified semantic space via a lightweight ViewAdapter, enabling coherent cross-view reasoning without incurring additional training overhead. A context-adaptive fusion module dynamically re-weights views to emphasize subtle yet semantically consistent threat signals, and the fused representation is integrated with task prompts for lightweight LLM fine-tuning. Experiments on CERT datasets show that Insight-LLM improves F1 by up to 4.8% and reduces false positives by 61%, while decreasing training time per newly added view by up to 83.2% compared with the simple Ensemble method.