home
CFP & Dates
Program
|
The KDD 2011 Workshop on Data Mining Applications in Sustainability seeks high-quality
submissions on the application of data mining to address problems in sustainable
operation and management. The workshop seeks to bring together data mining
researchers with interest in sustainability applications with researchers in
sustainability who are keen to apply knowledge discovery solutions in their domains.
Topics of interest can be organized into both "horizontals" and "verticals".
Vertical domain areas include (but are not limited to):
- Information Technology (IT)
- How can data centers be holistically (considering power sources, cooling
infrastructure, IT demand, performance) designed based on sustainability
metrics?
- How can we mine performance of IT systems (e.g., microprocessors, clusters)
to better understand opportunities for optimization of energy usage?
- Transportation
- How can fuel efficiency be improved through characterization of driver
behavior, vehicle operational patterns, etc. ?
- How can we design better transportation networks to account for drastic
scenarios (e.g., airport closures, winter storms across the eastern seaboard)
- How can environmental impact of transportation designs be minimized?
- Power Grid/Smart Grid
- What are the effects of changing demand (e.g., electric vehicles) and supply
(e.g., distributed generation) on the overall performance such as efficiency
and stability?
- How can data analytics techniques be applied to improve the management of
power/Smart Grid?
- How do we integrate distributed generation/micro-grids/energy storage?
- Water Infrastructure
- How can water supply and demand be optimally matched?
- How can data analytics techniques be applied to water infrastructure
management e.g., to detecting and/or predicting anomalies such as leaks?
- How can data analytics make the management of water infrastructure more
efficient?
- Building Energy Management
- How can per-appliance power consumption be determined without intrusive
instrumentation and in a cost-effective manner (the disaggregation problem)?
- How can appliance usage patterns be characterized, and leveraged to manage
power demand?
Horizontal aspects include but not limited to:
- Anomaly detection (for identifying sources of inefficiency)
- Causal diagnosis (to design system-wide models for fault detection and risk
assessment)
- Data fusion (for integrating heterogeneous, potentially correlated, data streams)
- Process modeling (for integrated understanding of equipment and infrastructure)
- Relationship discovery (for capturing associations between components)
- Mining massive throughput sensor streams (many data sources for sustainability
engineering are engineered through sensors, hence this is an important issue)
In addition, there are opportunities at the intersection of two or more verticals
(e.g., at the confluence
of water and power infrastructure). Such perspectives greatly aid in achieving end-toend
efficiency and appropriate energy/materials cost at all levels of infrastructure.
Paper Submission:
Two types of papers in ACM SIGKDD format are encouraged: long papers
with a maximum of 9 pages describing completed work on data
mining problems in sustainability and short papers with
a maximum of 6 pages describing ongoing research or preliminary results.
A PDF file should be emailed to the program chairs (naren@cs.vt.edu, manish.marwah@hp.com).
Novel work that straddles multiple facets of
data mining, modeling, and experimental
methodology are especially encouraged. The proceedings will be published
by ACM along with the main conference proceedings.
Important Dates:
- Submission: June 3, 2011
- Notification: June 16, 2011
- Camera ready versions due: June 21, 2011
- Workshop: Aug 21, 2011 (morning)
|