Accurate multivariate time series forecasting (MTSF) is critical for intelligent web services in Web of Things. When confronted with unseen multivariate time series (MTS), the industry typically invests significant time and resources in training multiple models to identify the optimal model for deployment. This paper proposes a novel, efficient, and scalable MTSF model selection method that directly selects suitable MTSF methods based on data characteristics without extensive model training. Model selection is a core component of AutoML, which has made significant progress in recent years. However, existing methods incur high operational costs and cannot be directly applied to MTSF tasks. Moreover, there is a lack of a comprehensive and cohesive public time series library for MTSF model selection. To address these challenges, we compile the first large heterogeneous labeled MTSF model selection dataset, called the ModelPile, which covers 41 mainstream datasets across 11 domains. We then propose AutoMTSF, a large model-enabled model selection method that transforms the MTSF model selection problem into a time series classification problem and utilizes the ModelPile to unlock large-scale multi-dataset training. AutoMTSF first uses the pre-trained large model to encode raw MTS. Given the coarse-grained limitations of large model encoding, Recursive Temporal Pattern Feature (RTPF) is proposed to capture both fine-grained and global temporal feature evolution, thereby effectively mapping data characteristics to the MTSF method space. Experiments comparing AutoMTSF with 2 baselines, 17 MTSF methods, and 4 large time series models show that AutoMTSF outperforms state-of-the-art methods while maintaining comparable execution time. This work represents a critical step in validating the accuracy and efficiency of large model-enabled classification for MTSF.

Machine learning models often inadvertently learn and perpetuate biases present in training data, particularly concerning sensitive attributes like gender and race. While existing fair representation learning approaches attempt to address this issue, they face challenges in balancing information preservation with bias elimination. This paper addresses the challenge of algorithmic bias in machine learning models by proposing a novel idempotence-constrained fair representation learning framework. We introduce a two-stage architecture that effectively eliminates both explicit and implicit bias while maintaining model performance. The first stage employs an idempotent encoder to remove explicit bias and enhance feature robustness through adaptive adversarial training, while the second stage utilizes a privacy filtering unit to eliminate implicit bias and task-irrelevant features. Our framework optimizes four key objectives: task performance, information preservation, sensitive information elimination, and representation stability. Theoretical analysis demonstrates convergence guarantees and representation stability of our approach. Extensive experiments on benchmark datasets, including UCI Adult and Heritage Health, show that our method achieves state-of-the-art performance in both fairness and accuracy metrics.

In the task of answering questions related to electricity knowledge, accurately understanding user intentions is fundamental to building an effective reasoning process. Questions in this domain often cover complex issues such as troubleshooting, electricity bill consultation, and electricity safety. User queries frequently involve multiple aspects and varying levels of information. Traditional single-modal methods struggle to fully address these diverse needs. Intention understanding plays a crucial role in constructing a clear and logical reasoning chain by accurately identifying the user’s core requirements. Through in-depth analysis of user intentions, the system can better organize and reason with multimodal information such as text, voice, and images—leading to logically coherent and accurate responses. Intention understanding not only enhances the accuracy of the reasoning process but also significantly improves the response efficiency and service quality of the electricity knowledge question-and-answer system. Current approaches to intention analysis primarily rely on classification-based methods, which limit the flexibility and richness of intent understanding. To broaden the scope and depth of user intent recognition, we introduce MMIntent-LLM, a novel framework that combines T5 with a Graph Transformer to generate structured intent representations from multimodal social media content. Our approach introduces three key innovations: (1) a structured intention reasoning framework based on ATOMIC, which provides a systematic method for decomposing intent generation into interpretable components; (2) a graph-based alignment mechanism for multimodal data (text and images) that ensures semantic consistency across modalities; and (3) an adaptive fine-tuning strategy that effectively transfers knowledge from pre-trained language models to the specific intent generation task. Extensive experiments on our multimodal social intention dataset show that MMIntent-LLM achieves state-of-the-art performance, improving the average BERT score by 8.7% and human evaluation scores by 12.3% compared to baseline methods.