We invite submissions to the 29th International Conference on Discovery Science, welcoming paper submissions from all areas relevant to artificial intelligence and data science.
For all information, please see: https://discovery-science.org/

The Discovery Science 2026 conference provides an open forum for intensive discussions and exchange of new ideas among researchers working in different areas of Artificial Intelligence and the data sciences, focusing on discovering and advancing scientific knowledge. Its scope includes developing and analyzing methods for discovering scientific knowledge, coming from machine learning, data mining, intelligent data analysis, and big data analytics, as well as their application in various domains of physical, life, environmental, natural and social sciences. This year, we particularly invite submissions dealing with methodological contributions to the automation or partial automation of discovery in various domains of the natural and human sciences, from a methodological perspective.

Discovery Science 2026 is part of the AI4Science Week and co-located with the 2nd International Conference on Artificial Intelligence for Science (AI4Sci 2026), which features separate AI4Physics, AI4Chemistry&Materials, AI4LifeSciences, and AI4Humanities&SocialSciences tracks dealing with applications of AI and Machine Learning. You are welcome to submit abstracts in any of those tracks as well, especially if your contribution focuses on the application of existing techniques to a specific field of scientific inquiry. Contrary to the main track of Discovery Science 2026, note that AI4Sci tracks are non-archival and also suited to ongoing or already published work, similarly to workshops co-located with other major conferences.
Aims & Scope

The current call for papers is tentative, and will be updated during the year. This year’s conference focuses on the following topics:

* Automated and semi-automated applications of Machine Learning and Artificial
* Intelligence in the discovery of scientific knowledge
* Benchmarking methodologies for the discovery of scientific knowledge
* Agentic AI methods and applications in scientific discovery
* Retrieval-Augmented Generation methods for the automation, both partial and full, of the scientific process

* Hypothesis generation methodologies, both causal and observational
* Causal discovery for the scientific process
* Methods for the integration of Machine Learning and AI in scientific laboratories, both at the hardware and software level
* Computational equation discovery and Symbolic Regression
* Methods and applications for safe and trustworthy AI (e.g. private, fair, transparent, sustainable, efficient) in the sciences, esp. social sciences and humanities

While the intended emphasis of the conference is on the above topics, we also value and welcome contributions dealing with more general Machine Learning, Data Mining and AI topics:

* Machine Learning, including supervised learning, unsupervised learning, semi-supervised learning, self-supervised learning
* Active learning, online learning, transfer learning, continual learning etc.
* Reinforcement learning
* AutoML, Meta-Learning, Planning to Learn
* Representation learning for vision, text, audio, language, and other data modalities
* Knowledge Discovery and Data Mining
* Anomaly and Outlier Detection
* Causal Modeling and reasoning
* Neuro-symbolic learning & hybrid AI systems (logic & formal reasoning, etc.)
* Physics-informed machine learning
* Data and Knowledge Visualization
* Explainable AI and Interpretable Machine Learning
* Human-Machine Interaction for Knowledge Discovery and Management
* AI and High-performance Computing, Grid and Cloud Computing
* Optimisation
* AI Creativity

Learning from Complex Data

* Data Streams, Evolving Data, Change Detection & Concept drift
* Time-Series Analysis
* Spatial, Temporal and Spatio-temporal Data Analysis
* Unstructured Data Analysis (textual and web data)
* Learning on graphs and topological deep learning
* Complex Network Analysis
* Process Discovery and Analysis
* Evaluation of Models and Predictions in Discovery Setting
* Applications of the above techniques in scientific domains, such as physical sciences (e.g., materials sciences, particle physics), life sciences (e.g., biology, medicine, neuroscience etc.), environmental sciences, natural and social sciences.

Submission Procedure

Contributions, written in English, must be formatted according to the guidelines of the Lecture Notes of Computer Science (LNCS) series by Springer-Verlag, which are available together with templates here.

DS 2026 will use The Microsoft CMT service to manage the peer-reviewing process for this conference. The Microsoft CMT service was used for managing the peer-reviewing process for this conference. This service was provided for free by Microsoft and they bore all expenses, including costs for Azure cloud services as well as for software development and support.

Please submit your contributions at https://cmt3.research.microsoft.com/DS2026/

Preprint Policy
Papers that have been published on preprint servers, such as arXiv, are eligible for submission.

Main conference
We invite original contributions that have not been previously published to another venue. The page limit for the contributions submitted to the main conference, including special sessions, is 15 pages in the LNCS format (note that 15 pages in LNCS format is roughly equivalent to 8 pages in NeurIPS format). This includes figures, title pages, references, and appendices.
Submission of the camera-ready version of the paper has to include the authors’ consent to publish on the above Springer LNCS website. Submissions will be reviewed following a so-called “single-blind” procedure. Therefore, authors should include their names and affiliations in the paper, but reviewers will be anonymous. Authors may not submit any paper that is under review elsewhere or has been accepted for publication in a journal or another conference; neither will they submit their papers elsewhere during the review period of DS 2026.
Accepted papers will be published in the Discovery Science Springer LNCS series.