Overview

CAESAR: Causal Anomalies - Exploring Synergies of Anomaly Detection and Causal Inference is a workshop held at ECML PKDD 2026, 07/09/2026 - 11/09/2026 in Naples, Italy.

This half-day workshop brings together researchers and experts from machine learning, data mining, and statistics, especially in the areas of causal inference, anomaly detection and their intersection, to share knowledge, discuss challenges, and progress in bringing these research areas together. Anomaly detection aims at identifying unusual or unexpected patterns that deviate from some, often unknown, normal state. The objective of causal inference is to identify the underlying causal relationships within data. The combination of these two concepts offers a more holistic comprehension of the issues we encounter and the systems we examine. One approach to combine both research fields is root cause analysis, which employs causal inference techniques to elucidate the source of observed anomalous behavior. Tracing back the cause of anomalous behavior is of particular importance in applications that monitor critical infrastructure, such as IT networks or power grids. Other approaches target the identification of anomalous causal relations with the objective of detecting critical changes in the data-generating process. The intersection of causal inference and anomaly detection presents a number of special challenges, as it requires the analysis of large amounts of data in order to learn the underlying causal relations. However, the availability of such amounts of data is often limited when dealing with anomalies that are notoriously rare. The objective of this workshop is to facilitate an exchange of ideas and foster collaboration between researchers, experts, and practitioners from both fields, as well as their intersection. The aim is to identify challenges and solutions, and to build a community of practice around these fields.

Topics of interest

Call for Papers

The workshop welcomes contributions (extended abstracts, 2-4 pages, and full papers, 6-16 pages, all in LNCS format), that cover, but are not limited to, one or several of the topics of interest.

Submissions will be double-blind (anonymised) and reviewed by at least 2 program committee members.

We welcome explicitly submissions with a focus on specific applications and use cases. We also welcome oral-only presentations of already published works and interesting problem statements or conceptual ideas which are still work-in-progress. For this, please submit extended abstracts.

Authors that would not want their papers to apply for possible oral presentation should inform the organisers at the time of submission. Each accepted paper will be invited to propose a camera ready version of their article taking into account the reviewers recommendations. Note that being accepted as a poster does not require reducing the length of the article. The organizers prepare the list of oral presentations, considering the program constraints and the scientific interest for broader exposition of the work.

The Workshop will be included in a joint ECML PKDD Post-Workshop proceeding published by Springer Communications in Computer and Information Science, in 1-2 volumes, organized by focused scope. Authors will have the faculty to opt-in or opt-out. For more info see here.

At least one of the authors must be registered to the conference. If not, the paper will not appear in the program.

Submission details

Submit your paper here!

Important Dates

Milestone Date
Abstract Submission deadline 05.06.2026
Paper Submission deadline 12.06.2026
Acceptance notification 20.07.2026
Camera-ready deadline 21.08.2026
Workshop day 11.09.2026

* all deadlines expire on 23:59 AoE

Keynote Speaker

We are delighted to announce that Charles Assaad will be giving the keynote talk.

Charles Assaad

Charles Assaad is a researcher whose work focuses on causal discovery, causal reasoning, and root cause analysis, with applications ranging from IT monitoring to epidemiology and public health. He leads the CIPHOD team at the Pierre Louis Institute of Epidemiology and Public Health (Sorbonne Université, Inserm), which develops causal inference methods for public health using large observational health databases. His recent research includes targeted causal discovery, causal abstractions such as summary causal graphs, cluster graphs and difference graphs, and methodological advances at the interface of causal inference, epidemiology, and public health.

Program

The workshop program will be announced soon.

Organizers

Program Committee

We thank all program committee members for their contribution to a fair and high quality review process.

Contact

Help us!

Since this is the first edition of the CAESAR Workshop, we would appreciate any help in advertising the Call for Papers. Feel free to distribute the website and flyer in your network.

Download Flyer