Photovoltaic (PV) power forecasting is often developed under an implicit assumption of stationarity, yet the weather–power relationship evolves over time and induces distribution shifts commonly known as concept drift. Existing PV power forecasting models either ignore this issue entirely or address it in a limited and insufficient manner. To better understand this non-stationarity, we distinguish between internal and external drift, which motivate the key components of our design. Therefore, we present ADrift , a drift-aware forecasting framework that integrates patch-based temporal modeling, weather-guided representation learning, and lightweight online adaptation. To model internal drift, the backbone employs prototype-guided weather experts that capture diverse meteorological patterns within each input window. To cope with external drift, a learnable adapter updates its parameters through a temporal gap attention mechanism that enables targeted adjustments to the model. In addition, a proactive update strategy further mitigates supervision delays under rapidly changing conditions. Experiments on three real-world PV datasets show that ADrift consistently improves forecasting accuracy over static and online-learning baselines, demonstrating its potential for practical deployment under evolving weather conditions.

• To the best of our knowledge, this work is the first to quantify weather-type imbalance in PV power datasets based on the intrinsic volatility of PV power, rather than relying on external parameters. • We observe that samples with high PV power volatility account for most of the training loss, which significantly reduces forecasting performance. • We design a novel volatility-aware re-weighting framework (ReMAV) that adjusts the importance of training samples based on their volatility levels, thereby improving model accuracy under imbalanced PV power datasets. • We validate the proposed framework on three benchmark datasets and demonstrate that our proposed ReMAV framework effectively handles weather-type imbalance in PV power datasets and consistently outperforms existing baseline models in forecasting accuracy. Recent short-term photovoltaic (PV) power forecasting methods have primarily focused on improving model architectures to enhance forecasting accuracy, they often overlook the issue of weather-type imbalance in PV power datasets. To this end, we first introduce a new metric, Mean Accumulated Volatility (MAV) , which quantifies the volatility of each sample. By translating unquantified weather-type imbalance into a measurable form of volatility imbalance, we observe that high-MAV samples account for most of the training loss, thereby harming the model’s forecasting accuracy. Then, we further propose ReMAV , a volatility-aware Re -weighting framework that down-weights the losses of high-MAV samples and up-weights those of low-MAV samples based on the MAV -based density. Extensive experiments on eleven baseline forecasting models across three real-world PV power datasets demonstrate that our proposed ReMAV framework effectively handles PV power with weather-type imbalance and consistently outperforms existing baseline models in forecasting accuracy. For example, on the Alice Springs dataset, ReMAV reduces average MAE by 8.53% over baselines, while on the PVOD dataset, MAE drops by 5.46% on average.

Fanyi Yang, Jianfeng Liu, Xin Zhang, Haoyu Liu, Xixin Cao, Yuefeng Zhan, Hao Sun, Weiwei Deng, Feng Sun, Qi Zhang. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.

Polarized Communities Search (PCS) aims to identify query-dependent communities where positive links predominantly connect nodes within each community, while negative links primarily connect nodes across different communities. Existing solutions primarily focus on modeling network topology, disregarding the crucial factor of node attributes. However, it is non-trivial to incorporate node attributes into PCS. In this paper, we propose a novel method called CO-guided RAndom walk in attributed signed networks (CORA) for PCS. Our approach involves constructing an attribute-based signed network to represent the auxiliary relations between nodes. We introduce a weight assignment mechanism to assess the reliability of edges in the signed network. Then, we design a co-guided random walk scheme that operates on two signed networks to model the connections between network topology and node attributes, thereby enhancing the search outcomes. Finally, we identify polarized communities using the Rayleigh quotient in the signed network. Extensive experiments conducted on three public datasets demonstrate the superior performance of CORA compared to state-of-the-art baselines for polarized communities search.

Knowledge Tracing (KT) aims to predict future students’ performance via their responses to a sequence of questions, which serves as a fundamental task for intelligent education. Most of the existing efforts directly predict students’ performance depending on their dynamically changing knowledge states . However, the individualization of questions is neglected and difficulty level differ from question to question, which would give some valuable clues to KT. Towards this end, in this paper, we propose a novel Multiple Question Factors for Knowledge Tracing (MQFKT) method, which fully exploits various question factors to generate better prediction. On one hand, calibrated student-concept connection space is established to obtain fine-grained response representations on questions according to the information of responses on questions. On the other hand, individualized difficulty levels with particular concept for different questions are introduced for improving the prediction performance. Extensive experiments on three datasets have shown that the MQFKT approach achieves more precise prediction of student performance and better interpretation of the model.

Polarized communities search aims at locating query-dependent communities, in which mostly nodes within each community form intensive positive connections, while mostly nodes across two communities are connected by negative links. Current approaches towards polarized communities search typically model the network topology, while the key factor of node, i.e., the attributes, are largely ignored. Existing studies have shown that community formation is strongly influenced by node attributes and the formation of communities are determined by both network topology and node attributes simultaneously. However, it is nontrivial to incorporate node attributes for polarized communities search. Firstly, it is hard to handle the heterogeneous information from node attributes. Secondly, it is difficult to model the complex relations between network topology and node attributes in identifying polarized communities. To address the above challenges, we propose a novel method Co-guided Random Walk in Attributed signed networks (CoRWA) for polarized communities search by equipping with reasonable attribute setting. For the first challenge, we devise an attribute-based signed network to model the auxiliary relation between nodes and a weight assignment mechanism is designed to measure the reliability of the edges in the signed network. As to the second challenge, a co-guided random walk scheme in two signed networks is designed to explicitly model the relations between topology-based signed network and attribute-based signed network so as to enhance the search result of each other. Finally, we can identify polarized communities by a well-designed Rayleigh quotient in the signed network. Extensive experiments on three real-world datasets demonstrate the effectiveness of the proposed CoRWA. Further analysis reveals the significance of node attributes for polarized communities search.