[Announcements] 2nd CALL FOR PAPERS - International Workshop on Anomaly and Novelty driven Self-Organizing Sensor Data Systems (AnNoSense)

Fri May 6 03:19:26 EDT 2022

[Sorry for cross-postings]

------------------------------------------------------------------
Dear researchers,
please consider submitting your works to AnNoSense 2022, co-
located with EWSN 2022, Linz, Austria.

AnNoSense 2022 –
Workshop on Anomaly and Novelty driven Self-Organizing Sensor Data
Systems

Website: https://sites.google.com/view/annosense/
------------------------------------------------------------------

AnNoSense CFP
We are all surrounded by millions of embedded systems that collect, process,
and communicate data about us. The pure amount of data captured by
sensors makes it nearly impossible to understand nor to model the
fundamental and highly interwoven knowledge captured in this data.

As one of the fundamental principles in computer science, Abstraction will
play a significant role in understanding, processing, and handling the data on a
large scale. Not in all situations, there is the need to have a comprehensive
understanding of the data on the lowest level (i.e., fine-grained activities of
daily living, interaction of employees with machines, etc.). Impacting
information is more on the level of (i) understanding changes in the data and
(ii) detecting the appearance of novel artifacts. Future systems must have the
ability to detect them and react accordingly to establish models for a lifelong
learning and adaptation process. Any static defined system that cannot cope
autonomously with changes and drifts occurring in the applied application
scenario will fail. Super- and semi-supervised methods that heavily rely on
expert knowledge and training are the first step but must pave the way to
systems capable of adapting and changing to occurring novelties to get
around training data limitation and insufficiency.

AnNoSense asks questions on the potential and opportunities of Recognizing
the Unknown – What is behind the Black Box of Big- and DeepData. Detecting
and predicting outcomes is helpful, but the challenge arises to explain, model,
and describe the cause- and effect relationships from theoretical and practical
perspectives.
DeepData leads the path to combine data with the theory to understand and
gain insights beyond what is already known. We hypothesize that, in contrast
to state-of-the-art machine learning algorithms, novelty and anomaly-driven
adaptive models in combination with semantic graphs that react to novel
occurring events and changes are more suitable for lifelong adaption and
understanding. Identifying something as new or anomalous allows drawing
conclusions and extending the experience captured in a model dynamically
and by far over the originally captured limited knowledge present in the initial
training data.

At the same time, the increasing complexity of such models also poses
challenges for understanding them, especially from the perspective of end-
users. The opacity of machine learning models has evoked new movements
and research domains, which argue that AI decision systems must be able to
explain their behavior to all involved stakeholders transparently. The
European Parliament states that AI systems should be “understandable to
non-technical audiences [..] which is necessary to evaluate fairness and gain
trust''. In this regard, new methods must be developed that to make complex
sensor-based “systems-of-systems” more transparent, but also include non-
experts in the design and training process. This will be necessary to gain
individual and societal trust and acceptance.

AnNoSense addresses the following foundational research concerns and
suggests topics including (but are not limited to):
- Understanding the impact of occurring novelties and anomalies in data
streams on established machine learning and recognition pipelines.
- Establishing Frameworks and Guidelines that go beyond classical label-based
training and understanding towards an open world assumption where neither
the infrastructure, the labels itself, nor the data is known in advance. These
systems must be based on a deep autonomous identification of various cause-
and effect relations to understand the underlying complex system.
- Laying down new foundations for the modeling, detecting, and
understanding the inclusion of anomalies and novel appearing artifacts in the
design and training of dynamic, adaptable machine learning models.
- Identifying Models and Tools on a theoretical and practical base, applicable
for establishing autonomous and dynamic system behavior based on
identified changes in the collected data streams.
- Discuss new Software Engineering Perspectives on model integration in data-
driven workflows and application scenarios, concerning real-time behavior,
scalability, bootstrapping, dynamic learning, failure and error prevention and
recovery, and focusing on building users' trust in the emerging systems.s
- Dynamic model training, adaption, and replacement guided by reasoning
processes dependent on detected novelties inducing an autonomous lifelong
learning and adaptation process.
- Semantic Modelling of heterogeneous data for Collaborative Reasoning that
combines Machine Learning Models with fundamental Theory to identify
latent Emerging Effects that are traceable to their origin, resulting in an
explainable AI-based system
- Understanding and Dealing with the Impact of Systems that autonomously
reacts to identified anomalies and novelties and alters its operating
environment on an intervention-based level.
- Identify the need to make systems aware that they probably change
environments and behaviors based on their simple presence.
- Reliability and Trustworthiness have a significant impact on the user
acceptance of dynamically changing systems. Addressing and presenting
system changes to users in an understandable and explainable way is a major
issue that needs to be addressed to implement novelty-driven systems
successfully. This includes Ethics, Privacy, Accessibility, Fairness, and
Interactivity in the design process, especially if vulnerable groups (e.g.,
patients, children, minorities, etc.) are affected.

Full Papers
Regular paper submissions must present original, highly innovative,
prospective, and forward-looking research in one or more of the themes given
above. Full papers must break new ground, present new insight, deliver a
significant research contribution, and provide validated support for its results
and conclusions. The workshop solicits
(i) conceptual papers describing proposals for novel methodologies, theories,
and principles that might be used to design, develop and build, analyze and
operate anomaly and novelty driven self-organizing autonomous sensor data
systems, (ii) observational, epistemological, and user study papers to deliver
evidence for possible future scenarios, and emerging platforms and
technologies as well as (iii) system- development papers proposing innovative,
novel HW/SW platforms.

Submission
Each paper must be submitted as a single PDF file in the EWSN 2022 format
(no longer than six pages in length) using the workshop paper submission
system on the workshop webpage. Submissions to this workshop must not be
under review by any other conference or publication during the workshop
review cycle and must not be previously published or accepted for publication
elsewhere.

Latex Templates are available at:
http://ewsn2022.jku.at/wp-content/uploads/2022/02/ewsn-template.zip

Reviewing Process
The selection of workshop participants will be carried out through a peer-
review process. Experts from related research fields will serve as reviewers to
guarantee fair decisions. Submissions need not be anonymous; however,
reviews will be realized anonymously using the evaluation form provided by
the submission system. Please refer to the paper submission link at the
workshop website (https://sites.google.com/view/AnNoSense). Questions
about papers should be directed to the Workshop chairs.

Publishing
Accepted papers will be included in the EWSN 2022 adjunct proceedings.

Workshop Chairs
Gerold Hoelzl (University of Passau, Germany)
Sebastian Soller (Almanara Research, Germany)
Philipp Wintersberger (Technical University Vienna, Austria)

IMPORTANT DATES
Submission Deadline June 1st, 2022
Notification of Acceptance July 1st, 2022
Camera Ready Deadline August 8th, 2022
Workshop October 3rd, 2022

best regards
Gerold Hoelzl

--
Dipl.-Ing. Dr. techn. Gerold Hoelzl

Chair of AnNoSense2022 WS (sites.google.com/view/annosense<http://sites.google.com/view/annosense>)

Embedded Interactive Systems Laboratory
Faculty of Computer Science and Mathematics, University of Passau
IT-Center / International House, Innstraße 43, 94032, Passau

Email:  gerold.hoelzl at uni-passau.de<mailto:gerold.hoelzl at uni-passau.de>
Web: http://www.eislab.net/team/gerold-hoelzl/
LinkedIn: http://lnkd.in/5NV83W

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ubicomp.org/pipermail/announcements_ubicomp.org/attachments/20220506/3729fd35/attachment.htm>