Multivariate time series forecasting is fundamental to web-scale systems. However, frequency-domain forecasters face two structural challenges: (i) frequency bias, where low-amplitude yet informative temporal cues are often submerged in noise and neglected during training; (ii) spectral degradation, where standard neural transformations distort high-amplitude periodic structures, thereby weakening the predictive signal. To address both issues, we propose GFMixer, a decoupled dual-path architecture. GFMixer first applies a Temporal Gradient Block (TGB) to capture low-amplitude information through adaptive selection based on temporal-geometric evidence. Subsequently, a Fourier-Aware Attention Block (FAB) represents high-amplitude, multi-frequency information to mitigate spectral degradation. Finally, both information streams are integrated via time-aligned residual connections within a Gradient Aggregation Block (GAB) for the final forecasting task. Extensive evaluations across seven standard benchmarks (ETT, Weather, Electricity, and Traffic) demonstrate GFMixer's superiority, securing 37 first-place rankings and the lowest average MSE overall. Beyond standalone performance, GFMixer serves as a versatile plug-and-play module that consistently enhances mainstream backbones. Code is available at: https://github.com/superlin30/GFMixer.

In the realm of deep graph learning, our study uniquely addresses the under-explored area of vague graph learning. While the effectiveness of deep graph learning is recognized across various disciplines, the nuances of vague graph learning — whether its inherent vagueness should be incorporated or disregarded and its influence on deep graph learning efficiency — remain largely unexamined. We fill this gap by introducing a novel decoupled graph learning framework. This is achieved by proposing a matrix-based or a tensor-based fusion module to estimate unobservable node attributes, a hybrid attention mechanism to bridge nodes with both explicit and implicit relationships, and a message-passing mechanism for feature-sensitive transporting. The design principle of decoupling allows it to accommodate ambiguities in any or all of these aspects of node representation, linking, and message passing. Furthermore, we leverage an extensive stock dataset spanning 64 years across the entire U.S. market to assess our framework. This real-world data not only adds a practical dimension to our study but also highlights the effectiveness of vague graph learning. Remarkably, our framework demonstrates superiority over state-of-the-art algorithms, marking performance enhancements of at least 6.73%, 7.25%, and 11.39% in terms of Rank IC,$R^{2}$, and Rank ICIR, respectively.

The relation between corporate environmental, social, and governance (ESG) performance and firm value has received increasing attention. However, the recent literature on corporate ESG performance evaluation suffers from potentially biased data sources, incomplete coverage, and uncertainty among agencies. The objectivity, timeliness, and breadth of news allow the media to provide a good perspective on ESG instead. This study investigated the effects of ESG-related news on the stock market to confirm that news is a helpful complement to the assessment of ESG performance. The results show that 1) ESG news has a greater and more significant impact on the stock market than news unrelated to ESG. 2) Of the three fields of ESG news, environmental news has a greater and more significant impact, which varies widely by topic. 3) Environmental news has different impacts on listed firms in different regions. This study provides new ideas for a complete and objective assessment of corporate ESG performance and the management of a corporate media image from the use of ESG news.