CALL FOR SPECIAL SESSION PAPERS: Time Series Methods for Dynamic Knowledge Fusion and Predictive Analytics

The KSEM 2025 Conference invites submissions of high-quality research papers to the Special Session on Time Series Methods for Dynamic Knowledge Fusion and Predictive Analytics. This session aims to explore the forefront of temporal data analytics and its integration into knowledge management frameworks, emphasizing innovative methodologies that drive predictive analytics in complex, real-world systems. As organizations increasingly rely on real-time data and sophisticated forecasting techniques to support strategic decision-making, this session seeks to advance research on the extraction, integration, and utilization of dynamic knowledge from time series data.

PAPER SUBMISSIONS: https://easychair.org/conferences?conf=ksem2025 (choose Special Session track)

We welcome submissions from academia, government, and industry that present novel research on both theoretical advancements and practical implementations in time series analysis with applications to knowledge fusion and predictive analytics. We encourage papers that contribute to the development, evaluation, and application of cutting-edge time series methodologies and their integration with knowledge-driven systems. Submissions may address, but are not limited to, the following topics:

• Advanced Time Series Modeling and Forecasting:

• Development and refinement of statistical models (e.g., ARIMA, VAR, GARCH) and machine learning approaches (e.g., LSTM, CNN, transformer-based models) for accurate forecasting in high-dimensional, nonstationary, and complex datasets.

• Hybrid and ensemble models that combine traditional time series methods with deep learning techniques to improve prediction accuracy and robustness.

• Innovative methods for handling irregularly sampled or missing data in time series, and for addressing nonlinearity and volatility in temporal patterns.

• Dynamic Knowledge Fusion:

• Techniques for integrating heterogeneous time series data from multiple sources, including multimodal data streams, sensor networks, and knowledge graphs.

• Algorithms and frameworks for the real-time fusion of dynamic data to support comprehensive knowledge representation and reasoning.

• Case studies on how integrated temporal data enhances decision-making processes in fields such as industrial monitoring, urban management, and healthcare diagnostics.

• Predictive Analytics for Intelligent Decision Support:

• Applications of time series forecasting in developing predictive analytics platforms that drive automated decision-making in domains such as cybersecurity, supply chain management, digital twins, and AI security.

• Studies on how time series insights can be utilized to optimize resource allocation, mitigate risks, and enhance operational efficiency.

• Comparative evaluations of predictive models in various industrial contexts, emphasizing the benefits and limitations of different approaches.

• Uncertainty Quantification and Interpretability:

• Methods for quantifying uncertainty in time series predictions, including Bayesian approaches, bootstrap methods, and ensemble techniques.

• Research on improving model interpretability and transparency, facilitating better understanding and trust in data-driven decision systems.

• Techniques for integrating uncertainty measures into knowledge fusion frameworks to support robust decision-making.

• Integration with AI and Knowledge Management Systems:

• Frameworks for combining time series analytics with large language models (LLMs), semantic web technologies, and knowledge graphs to enhance knowledge extraction and representation.

• Approaches to embedding temporal insights into AI-driven knowledge management systems, ensuring that predictive analytics contributes meaningfully to strategic planning and policy formulation.

• Investigations into the challenges of merging time series methods with knowledge-based systems, with emphasis on scalability, data quality, and real-time processing.

Submissions should follow the KSEM Conference paper format and be prepared in accordance with the standard Paper Submission Guidelines. Accepted papers will be published by Springer and submitted to the EI database. Excellent papers with extensions will be recommended for special issues in renowned journals.

SPECIAL SESSION CHAIR

• Ziyu Jia

Institute of Automation, Chinese Academy of Sciences, Beijing, China

• Xinliang Zhou

Nanyang Technological University, Singapore

• Idris Elbakri

Kyrgyz National University, Faculty of Informational and Innovational Technology, Kyrgyz

• Hairong Chen

Beijing Jiaotong University, Beijing, China