Time series data has become ubiquitous across numerous web systems, including web services,
cloud computing, IoT devices, digital health platforms, wearable technology, financial markets,
biological sciences, and environmental sciences. With the availability of massive amounts of data,
complex underlying structures & distributions, and advanced high-performance computing platforms,
there is a growing demand for new theories and algorithms to address fundamental challenges
(e.g., representation, prediction, generation, and causal analysis) across a variety of applications.
Time series analysis now plays a crucial role in addressing key challenges in web-based environments,
including server load balancing, anomaly detection in e-commerce traffic, and tourism demand
forecasting. This workshop aims to provide a platform for researchers and AI practitioners from
both academia and industry to discuss potential research directions, key technical issues, and
present solutions to tackle related challenges in practical web-based applications involving
time series data. We invite researchers, AI practitioners, and policymakers from the related
areas of machine learning, data science, statistics, econometrics, and many others to contribute
to this workshop.
This workshop encourages submissions of innovative solutions for a broad range
of web-centric time series analysis problems. Topics of interest include but are not limited
to the following:
Submissions should be 4-8 pages long, excluding references, and must follow the new standard ACM conference proceedings template. The review process is double-blind, so authors should ensure their identities are not revealed to reviewers. An optional appendix of arbitrary length is allowed and should be put at the end of the paper (after references). All manuscripts should be submitted in a single PDF file including all content, figures, tables, references, and other information.
For LaTeX users: unzip acmart.zip, make, and use sample-sigconf.tex as a template; Additional information about formatting and style files is available online at: https://www.acm.org/publications/proceedings-template
We will use EasyChair to manage the submission and peer-reviewing process for this workshop. Accepted papers will be presented as posters during the workshop and listed on the website (non-archival/without proceedings). A small number of accepted papers will be selected to be presented as contributed talks (15-minute oral presentations). We also welcome submissions of unpublished papers, including those submitted/accepted to other venues if that other venue allows.
Submission site: https://easychair.org/my/conference?conf=ai4ts25
Dual submission policy: This workshop is non-archival; all accepted papers will be available on this website, there are no formally-published proceedings. If a paper is currently under review at another venue, it can still be submitted to this workshop. If a paper has previously appeared in a journal, workshop, or conference, it should be reasonably extended in order to be accepted at this workshop. Parallel submission of papers under review at WWW 2025 is allowed.
!! TheWebConf 2025 fast-track submissions: Authors should include an Appendix following the references section, where they must directly paste the reviews received during the main review process. These reviews will be utilized to determine acceptance into the workshop. Modifying or altering reviewer comments is strictly prohibited. In the appendix, authors are required to include a section titled "Improvements", where they should briefly outline whether they were able to address any of the issues highlighted in the main review and provide details of the revisions made. Only minor revisions are permitted; substantial changes are not allowed. Additionally, it is entirely acceptable to report no revisions if the work is already in satisfactory condition. The provided reviews will not be shared with any third parties or individuals other than the workshop organizers. Strict adherence to privacy and the double-blind policy will be maintained.
Workshop Paper Submission Due Date:
18 December, 2024
1 January, 2025 (23:59pm AoE)
!! Fast-track Paper Submission Due Date: 26 January, 2025 (23:59pm AoE)
Notification of Paper Acceptance:
13 January, 2025
27 January, 2025
Workshop Paper Camera-ready:
2 February, 2025
7 February, 2025 (23:59pm AoE)
WWW'25 AI4TS Workshop Date: 29 April, 2025
Workshop Date: 29 April, 2025
Location: Room C4.10, ICC Sydney
Time (GMT+11) | Speaker | Title |
---|---|---|
9:00 am - 9:10 am | Dr. Yuxuan Liang Dr. Ming Jin | Opening Remarks |
9:10 am - 9:50 am | Prof. Longbing Cao | Keynote Talk 1: Irregular Multivariate Time Series Modeling: From Time to Frequency Domains |
9:50 am – 10:30 am | A/Prof. Ben Fulcher | Keynote Talk 2: Extracting interpretable structure from time series datasets using highly comparative time series analysis |
10:30 am – 11:00 am | - | Coffee Break |
11:00 am - 11:40 am | Dr. Chang Xu | Keynote Talk 3: Time Series Diffusion Models with Guided Learning Process: Strategies, Methods, and Real-World Applications |
11:50 am - 12:30 pm | Prof. Flora Salim Dr. Yuxuan Liang Dr. Ming Jin A/Prof. Ben Fulcher Dr. Chang Xu | Panel Discussion (Topic: Foundation & generation models, agentic reasoning systems, and general time series intelligence) |
12:30 pm - 1:30 pm | - | Lunch Break |
1:30 pm - 2: 10 pm | Prof. Bin Yang | Keynote Talk 4: Data Driven Decision Making with Time Series |
2:10 pm - 2: 50 pm | Prof. Geoff Webb | Keynote Talk 5: Convolutional Kernels for Effective and Scalable Time Series Analytics |
2:50 pm - 3:30 pm | - | Coffee Break |
3:30 pm - 4:30 pm | Oral Presentations | (Part A): Controllable Financial Market Generation with Diffusion Guided Meta Agent (Part B): Diffusion Graph Model for Time Series Anomaly Detection via Anomaly-aware Graph Sparsification and Augmentation |
4:30 pm - 5:00 pm | - | Poster Session |
Real-life multivariate time series (MTS) are often irregular, presenting irregularities including non-IID, stylistic, asymmetric, inconsistent, and dynamic characteristics. High-dimensional and multi-spectral multivariates are even more challenging to model. This talk reviews such challenges and briefly introduces some of our recent progress in deep MTS modelling of such irregularities, and non-IIDnesses (interactions, couplings, and heterogeneities) in time, frequency, or time+frequency domain. The approaches synergise deep neural learning with variational learning, copula methods, shallow-to-deep non-IID learning, and basis functions, etc.
Time series classification is a fundamental data science task, interpreting dynamic processes as they evolve over time. Convolutional kernels provide an effective method for extracting a wide range of different forms of information from time series data. I present the Rocket family of time series classification technologies that utilize convolutional kernels to achieve state-of-the-art accuracy with many orders of magnitude greater efficiency and scalability than any alternative. These make time series classification feasible at hitherto unattainable scale. The methods also have potential application across many other forms of time series analysis, including extrinsic regression, clustering, anomaly detection, segmentation and forecasting.
Many systems in the world around us evolve through time and can be measured in the form of multivariate time series. In this talk I will introduce systematic tools that we’ve developed for quantifying the dynamical properties of these systems and the interactions they contain. We describe our approach as ‘highly comparative’, as large numbers of possible analysis methods (e.g., >7000 time-series features implemented in our hctsa software package, and ~150 pairwise dependence measures in pyspi) are compared. Our approach underpins systematic ways of analyzing time-series data, leveraging an interdisciplinary theoretical literature and yielding fast and scalable analytics that are also interpretable. I will focus on the open tools and their software implementations that enable these analyses and summarize some recent applications of the approach, including to neuroscience datasets.
Time series data captures properties that change over time. Such data occurs widely, ranging from the scientific and medical domains to the industrial and environmental domains. As part of the continued digitization of processes throughout society, increasingly large volumes of time series are available. In this talk, we focus on data-driven decision making with time series data, e.g., enabling greener and more efficient transportation based on traffic time series forecasting. We introduce our research paradigm of "data-governance-analytics-decision." We first introduce the data foundation of time series, which is often multi-modal and heterogeneous. Next, we discuss data governance methods that aim to improve data quality. We then cover data analytics, focusing on five desired characteristics: automation, robustness, generality, explainability, and resource efficiency. We finally cover data-driven decision-making strategies and briefly discuss promising research directions.
Diffusion models have shown great potential in time series modeling by capturing complex temporal dependencies and generating realistic sequences through iterative denoising. In this talk, we explore innovative strategies that integrate various guidance mechanisms—targeted interventions that steer the model toward desired properties—into the diffusion process for both time series generation and forecasting. By incorporating different forms of guidance, such as text/embedding-based prompting, statistical constraints, and task-aware feedback, we enable flexible, personalized, and domain-aware time series generation. For forecasting, we leverage implicit guidance techniques to enhance stability, mitigate biases, and improve predictive accuracy. We will discuss the underlying principles, technical challenges, and practical benefits of these approaches, highlighting their potential to address real-world, cross-domain challenges.
* Diffusion Graph Model for Time Series Anomaly Detection via Anomaly-aware Graph Sparsification and Augmentation Disen Lan, Guibin Zhang, Rongjin Guo
* TimeDP: Learning to Generate Multi-Domain Time Series with Domain Prompts Yu-Hao Huang, Chang Xu, Yueying Wu, Wu-Jun Li, Jiang Bian
* Enhancing E-commerce Supply Chain Management through Large Time Series Model Shiyu Wang, Xinyue Zhong, Jiawei Li, Rongwei Liu, Yidong Feng, Congcong Hu, Fan Huang, Zhou Ye
* TSGGuide: Recommendation Guide for Multivariate Time Series Generation Beining Zhang, Kai Wu, Xiaoyu Zhang, Handing Wang, Jing Liu, Bei Wang, Kai Wang
* Controllable Financial Market Generation with Diffusion Guided Meta Agent Yu-Hao Huang, Chang Xu, Yang Liu, Weiqing Liu, Wu-Jun Li, Jiang Bian