Time-series forecasting is considered a critical task with extensive applications across diverse domains. However, effectively capturing both cross-dimension and cross-time dependencies in nonstationary time series remains a significant challenge, particularly due to the confounding effects of environmental factors. These factors often introduce spurious correlations that obscure the learning of meaningful temporal features. In this article, the novel framework Caformer (Causal Transformer) is proposed for time-series forecasting grounded in causal reasoning, the science of identifying causality. The framework consists of four key modules: the dynamic learner, environment learner, temporal learner, and decompose learner. The dynamic learner uncovers dynamic interactions among features to model cross-dimension dependencies, while the temporal learner infers cross-time dependencies under causal constraints. The environment learner, together with the decompose learner, extracts environmental factors and applies a backdoor adjustment to mitigate the confounding effects on the time series. Extensive experiments demonstrate that Caformer achieves the state-of-the-art performance in both long-term and short-term forecasting. It achieves up to a 26.2% mean square error (mse) reduction on the traffic dataset and 21.8% on the electricity dataset compared to PatchTST, with consistent gains across eight long-term benchmarks. On the M4 dataset, Caformer ranks first across all 15 short-term forecasting categories. In addition to strong predictive accuracy, Caformer provides interpretable insights into the learned dependencies.

The automated grading of subjective answers is crucial for reducing manual workload and enhancing feedback efficiency in online education, particularly for short answer scoring (SAS). However, in scenarios with limited labeled data, existing methods face challenges in sample diversity and scoring consistency due to limited training data and misaligned label distributions. While data augmentation and transfer learning have been employed to address these issues, rule-based approaches often lack textual variability, and general-domain embeddings struggle to align with domain-specific scoring criteria. To overcome these limitations, we propose the Scoring with Contextual Alignment and Language Enhancement (SCALE) framework, a novel LLM-driven training paradigm that synthesizes diverse responses while preserving scoring consistency. SCALE leverages a knowledge graph-based generation strategy to enhance sample diversity by substituting key phrases with contextually aligned alternatives and employs a style rewrite prompt to introduce linguistic variations. To mitigate label inconsistency, we introduce a polish align prompt that refines synthetic and real samples into a shared semantic subspace, training an annotator model for aligned scoring. Additionally, an entity-aware enhancement mechanism improves comprehension of formulas and quantitative content. Extensive experiments on multilingual and multi-domain datasets demonstrate that SCALE achieves state-of-the-art performance, improving Pearson scores by 4.9%, 2.27%, and 1.34% over BERT, RoBERTa, and ERNIE 3.0, respectively.

Due to the challenge posed by multi-source and heterogeneous data collected from diverse environments, causal relationships among features can exhibit variations influenced by different time spans, regions, or strategies. This diversity makes a single causal model inadequate for accurately representing complex causal relationships in all observational data, a crucial consideration in causal learning. To address this challenge, we introduce the nonlinear Causal Kernel Clustering method designed for heterogeneous subgroup causal learning, illuminating variations in causal relationships across diverse subgroups. It comprises two primary components. First, the construction of a sample mapping function forms the basis of the subsequent nonlinear causal kernel. This function assesses the differences in potential nonlinear causal relationships in various samples, supported by our causal identifiability theory. Second, a nonlinear causal kernel is proposed for clustering heterogeneous subgroups. Experimental results showcase the exceptional performance of our method in accurately identifying heterogeneous subgroups and effectively enhancing causal learning, leading to a great reduction in prediction error.