Description Workshop
Theme
The main goal of the ODD workshop is to bring together academics, industry and government researchers, and practitioners to discuss and reflect on outlier mining challenges. Outlier detection methods are often applied to numerous real-world applications like security, healthcare, finance. These tasks affect humans in some way, and hence, ensuring the fairness of such methods is paramount. Fairness relates to developing unbiased decision policies whose outcomes are not dependent on any sensitive features or variables such as gender and race. Transparency is another factor, linked to the fairness of methods, where the decision made by the designed methods should be understandable in order to ensure that the methods are not biased towards specific groups.
We want to highlight issues related to fairness and transparency and aim to increase awareness of the following topics:
- How can we measure the bias in the outlier detection methods?
- How can we ensure that outlier identification does not produce unjust outcomes for protected minority groups (such as age/race/sex)?
- How can we prevent a negative feedback loop, of unfair outlier detection based on historically biased data (such as over policing in minority neighborhoods)?
- How can we make outlier detection methods transparent?
- How can we employ deep learning models for detecting and ensuring fairness in outlier detection systems?
- How can we ensure statistical parity in our analysis, such that the outlier detection outcomes are independent of class memberships?
- How can we employ adversarial learning mechanisms for outlier detection to ensure fairness?
Schedule
Aug 15, 2021: Virtual (All times are Pacific Time)
All accepted papers will be presented in the Spotlight session.
| Morning Sessions | |
|---|---|
| 8:00 am | Opening Remarks |
8:10 am
|
Keynote: Sudipto Guha An Anomalous Talk on Anomaly Detection |
8:40 am
|
Keynote: Rajmonda Caceres Network Anomaly Discovery with Reinforcement Learning |
| 9:10 am | Invited Talk 1 Hongfu Liu Deep Clustering-based Fair Outlier Detection |
| 9:30 am | Break |
| 9:40 am |
FAIRNESS PANEL
|
| 11:10 am |
Spotlight Talks |
| 11:45 am |
Social Networking |
12:00 pm | Lunch Break |
| Afternoon Sessions | |
|---|---|
1:00 pm
|
Keynote: Leman Akoglu Fairness in Outlier Detection: Being Wise or Otherwise Slides |
1:30 pm
|
Keynote: Charu Aggarwal Ensemble-Centric Evaluation of Outlier Detection Slides |
| 2:00 pm | Invited Talk 2 Zhiwei Wang Multi-Scale One-Class Recurrent Neural Networks for Discrete Event Sequence Anomaly Detection |
| 2:20 pm | Break |
2:30 pm
|
Keynote: Danai Koutra Graph Summarization Meets Outlier Detection Slides |
3:00 pm
|
Keynote: James Verbus Preventing Abuse Using Unsupervised Outlier Detection Slides |
| 3:30 pm | Invited Talk 3 Huayi Zhang ELITE : Robust Deep Anomaly Detection with Meta Gradient |
| 3:50 pm | Closing Remarks |
Keynote Talks
Charu Aggarwal
IBM
Leman Akoglu
Carnegie Mellon University
Rajmonda Caceres
MIT
Sudipto Guha
Amazon
Danai Koutra
University of Michigan
James Verbus
Fairness Panel
Neil Shah
Snap
Moderator
Solon Barocas
Microsoft Research
Ian Davidson
UC Davis
Jing Gao
Purdue University
Deepak Padmanabhan
Queen's University Belfast
Hanghang Tong
UIUC
Accepted Papers
Choosing Effective Projections for Fast and Accurate Anomaly Detection
PDF
Chen Almagor, Yedid Hoshen
Anomaly Alignment Across Multiple Attributed Networks
PDF
Jie Zhang, Nannan Wu, Wenjun Wang, Ying Sun, Siddharth Bhatia
CSCAD: Correlation Structure-based Collective Anomaly Detection in Complex System
PDF
Huiling Qin, Xianyuan Zhan, Yu Zheng
Out-of-Distribution Detection and Fairness Assessment in Dermatology
PDF
Hannah H Kim, Girmaw Abebe Tadesse, Celia Cintas, Skyler D Speakman, Kush R Varshney
Anomaly Detection and Automated Labeling for Voter Registration File Changes
PDF
Sam F Royston, Courtenay Cotton
Scrutinizing Shipment Records To Thwart Illegal Timber Trade
PDF
Debanjan Datta, Sathappan Muthiah, John Simeone, Amelia Meadows, Naren Ramakrishnan
Scalable Change Point Detection for Dynamic Graphs
PDF
Shenyang Huang, Guillaume Rabusseau, Reihaneh Rabbany
The Effect of Hyperparameter Tuning on Comparative Evaluation of Anomaly Detection Methods
PDF
Jonas Soenen, Elia Van Wolputte, Lorenzo Perini, Vincent Vercruyssen, Wannes Meert, Jesse Davis, Hendrik Blockeel
Call for Papers
We welcome many kinds of papers, such as, but not limited to:
- Novel research papers
- Work that will be presented at KDD
- Demo papers
- Work-in-progress papers
- Papers on case studies of benchmark data
- Relevant work that has been previously published
While we aim for a focus on the theme of fairness and transparency, we welcome papers addressing any other challenges at large of the subject area. Topics of interest include, but are not limited to:
- Interleaved detection and description of outliers:
- Description models for given outliers
- Pattern and local information based outlier description
- Subspace outliers, feature selection, and space transformations
- Ensemble methods for anomaly detection and description
- Descriptive local outlier ranking
- Identification of outlier rules
- Finding intensional knowledge
- Contextual and community outliers
- Human-in-the-loop modeling and learning
- Visualization techniques for interactive exploration of outliers
- Comparative studies on outlier description
- Outlier mining for complex databases:
- Graph data (e.g. community outliers)
- Spatio-temporal data
- Time series and sequential data
- Online processing of stream data
- Scalability to high dimensional data
- Formal outlier mining models:
- Supervised, semi-supervised, and unsupervised models
- Statistical models
- Distance-based models
- Density-based models
- Spectral models
- Constraint-based models
- Ensemble models
- Applications of outlier detection and description:
- Fraud in financial data
- Intrusions in communication networks
- Sensor network analysis
- Social network analysis
- Health surveillance
- Customer profiling
- Related research fields:
- Contrast mining
- Change and novelty detection
- Causality analysis
- Frequent itemset mining
- Compression theory
- Subgroup mining
- Subspace learning
All papers will be peer reviewed and double-blinded.
Submissions must be in PDF, no more than 9 pages long (including references) — shorter papers are
welcome — and formatted according to the standard double-column ACM
Proceedings Style. Every effort must be made to preserve the anonymity of the authors. Authors may submit (optional) supplementary material, such as appendices, proofs, derivations, data, or source code; all supplementary materials must be in PDF or ZIP format. Like submissions, supplementary material must be anonymized. To submit supplementary material, first upload your submission. You will then be able to upload supplementary material from the author console. Looking at supplementary material is at the discretion of the reviewers.
The accepted papers will be published on the workshop’s website and will not be considered
archival for resubmission purposes. Authors whose papers are accepted to the workshop will have the opportunity to participate
in a poster session, and some set will also be chosen for oral presentation.
For paper submission, please proceed to the submission website.
Please send enquiries to siddharth@comp.nus.edu.sg
To receive updates about the current and future workshops and the Outlier Detection community, please fill your contact information, or follow on Twitter.
Important Dates
Paper Submission Deadline: May 20, 2021 June 1, 2021 (23:59 UTC-12)
Author Notification: June 30, 2021
Workshop: August 15, 2021
Workshop Organizers
Program Committee
Acar Tamersoy (NortonLifeLock)
Alan Fern (Oregon State University)
Aoqian Zhang (Beijing Institute of Technology)
Daochen Zha (Texas A&M University)
Deepak Padmanabhan (Queen's University)
Dhivya Eswaran (Amazon)
Evangelos Papalexakis (UC Riverside)
Eamonn Keogh (UC Riverside)
Fabrizio Angiulli (University of Calabria)
Feng Chen (UT Dallas)
Hongfu Liu (Brandeis University)
Jagdish Ramakrishnan (Facebook)
Jing Gao (Purdue University)
Jundong Li (University of Virginia)
Kaize Ding (Arizona State University)
Kijung Shin (KAIST)
Klemens Böhm (Karlsruhe Institute of Technology)
Lei Cao (MIT)
Lukas Ruff (TU Berlin)
Marius Kloft (TU Kaiserslautern)
Mihai Cucuringu (Oxford University)
Mohit Wadhwa (LinkedIn)
Mostafa Rahmani (Amazon Web Services)
Nannan Wu (Tianjin University)
Neil Shah (Snap)
Nesime Tatbul (Intel Labs and MIT)
Nikolay Laptev (Facebook)
Ninghao Liu (Texas A&M University)
Parikshit Gopalan (VMware Research)
Peng Gao (UC Berkeley)
Prateek Chanda (Microsoft Research)
Qingsong Wen (Alibaba)
Robert Vandermeulen (TU Berlin)
Ryan Rossi (Adobe)
Sambaran Bandyopadhyay (Amazon)
Samira Samadi (MPI-IS)
Shenghua Liu (Chinese Academy of Sciences)
Shirui Pan (Monash University)
Shubhomoy Das (Espressive)
Simon Woo (Sungkyunkwan University)
Sunil Aryal (Deakin University)
Susik Yoon (KAIST)
Tsuyoshi Ide (IBM)
Varun Chandola (SUNY, Buffalo)
Vincent Vercruyssen (KU Leuven)
Yedid Hoshen (Hebrew University of Jerusalem)
Yizhou Yan (Facebook)
Youcef Djenouri (SINTEF)
Yue Zhao (CMU)
Yuening Li (Texas A&M University)
Yujing Wang (Microsoft Research)
Zheng Li (Arima Inc.)