--------------------------------------------------------------------------
** Please feel free to forward to anyone who might be interested **
--------------------------------------------------------------------------


CALL FOR PAPERS
for the 1st International ECML/PKDD 2014 Workshop
on


*Machine Learning for Urban Sensor Data*
(SenseML 2014)


https://www.tk.informatik.tu-darmstadt.de/en/senseml-2014/
Sept. 15, 2014 - Nancy, France
*** Paper submission deadline extended: Friday, June 27th, 2014 ***
--------------------------------------------------------------------------

Over a decade, sensor research has proven the use of sensor networks for different use cases. The research has mostly focused on aspects of sensor network deployment, energy-efficiency, and wireless networking. Today, the focus is shifting from “How do we collect data?” to “What can we learn from the data and how do the models look like?” as more and more data becomes available. Especially urban applications – from people and car movement to building and environmental sensing – has been a constant driver of innovation over the past years, delivering an ever increasing amount of useful data. Even more data comes from participatory sensing as everyone collects data all the time. However, most of the data remains unused as the full potential of machine learning and data mining is yet to be utilized.

The challenges that are associated with urban sensor data for the design and application of learning algorithms are the focus of this workshop. Examples include algorithms that are suited to the special needs this kind of data imposes such as missing values, unreliable measurements, missing calibration or high spatial diversity. Also, the architecture of systems, i.e., the complete data analysis pipeline from the collection of measurements to the final model is of interest. New real-world data sets coming from these areas are also of special interest for the ML community.
Bringing together the sensor systems expertise from the Wireless Sensor Networks and Sensor Systems community with the knowledge available in the machine learning community, will open up a whole new set of applications and technical questions. SenseML will, therefore, promote work that on one hand concentrates on generating high quality data set, and on the other hand on applying or developing state of the art machine learning algorithms to yield highly accurate models. A special focus also lies on interpretable models such as rule sets or decision trees. Especially for sensor data, having interpretable models is of interest for discovering potential relationships in the sensor network. Proceeding this way, we expect to gain unique insights into sensor data and ideas for novel applications.

The workshop considers hot topics of both disciplines, position papers, novel ideas, in-progress work on system architecture (e.g., the data analysis pipeline), enabling technologies, and emerging applications.

TOPICS OF INTEREST
------------------
Topics of interest include, but are not limited to:

* Real-time machine learning
* Iterative machine learning
* Multi-target learning
* Generating data analysis pipelines
* Evaluation of machine learning models tailored to sensor data
* Data extraction from sensor networks
* Data conversion and calibration issues
* Meta-learning, e.g., learning to adjust the analysis pipeline automatically
* Interpretable models, e.g., Rule Learning or Decision Tree Learning
* Generating high-quality data sets
* Data quality issues
* Dealing with missing and low quality data
* Feature Engineering with a focus on sensor data features
* Feature weighting and combination
* Generating high-quality features from sensor data

SPRINGER BOOK
-------------
Depending on the number of submissions, it is planned to publish revised selected papers as a volume in the Springer LNCS/LNAI series.

WORKSHOP ORGANIZERS
-------------------
* Frederik Janssen, Knowledge Engineering Group, Technische Universität Darmstadt, Germany (janssen@ke.tu-darmstadt.de)
* Immanuel Schweizer, Telecooperation Group, Technische Universität Darmstadt, Germany (schweizer@cs.tu-darmstadt.de)

PROGRAM COMMITTEE
-----------------
* Johannes Fürnkranz (TU Darmstadt, Darmstadt, Germany)
* João Gama (LIAAD, Porto, Portugal)
* Kristian Kersting (TU Dortmund, Dortmund, Germany)
* Eneldo Loza Mencia (TU Darmstadt, Darmstadt, Germany)
* Benedikt Schmidt (TU Darmstadt, Darmstadt, Germany)
* Heiko Paulheim (University of Mannheim, Mannheim, Germany)
* Axel Schulz (SAP, Darmstadt, Germany)
* Christian Wirth (TU Darmstadt, Darmstadt, Germany)
* Nico Piatkowski (TU Dortmund, Dortmund, Germany)
* Florian Lemmerich (University Würzburg, Würzburg, Germany)

SUBMISSIONS AND STYLE
---------------------
We invite full papers (up to 16 pages) and short papers (up to 8 pages). Position papers or work-in-progress (4 pages) are also very welcome.
Depending on the accepted submissions, there will be a poster session additionally to the presentations.

The papers must be written in English and formatted according to the Springer LNAI guidelines, which can be downloaded at: http://www.springer.de/comp/lncs/authors.html.

Papers should be submitted via: https://www.easychair.org/conferences/?conf=senseml2014

If you have any further question please contact the SenseML Organizers. We recommend to follow the format guidelines of ECML/PKDD (Springer LNCS), as this will be the required format for accepted papers.

More details can be found on the workshop website:
https://www.tk.informatik.tu-darmstadt.de/en/senseml-2014/

IMPORTANT DATES
---------------
* Paper Submission Deadline (extended): Friday, June 27th, 2014
* Notification of Acceptance: Friday, July 11th, 2014
* Final Version: Friday, July 25th, 2014
* ECML conference: Monday, September 15th, 2014 - Friday, 19th September, 2014
* Workshop SenseML: Monday, September 15th, 2014