Workshop on Physical Analytics (June 16, 2014)

Despite the rapid rise in online activity, people lead much of their lives in the physical world. They travel to places, dwell at various locations, spend time with other people, shop, go to the gym, watch movies, and listen to music or are exposed to announcements, all of which is often based on what interests them. Therefore, there is much to be learnt about users from their physical actions and activities, which is often not manifested in their online activity. Such insights can be of benefit both to users directly and to businesses. For instance, a user could benefit from having a digital health diary automatically keep track of their food purchases and sleep patterns, or a digital personal assistant that melds together physical signals with online signals. A business such as a retail store could learn about the browsing habits of users while in their store, while an advertising network could stitch together information from user visits to multiple stores on a visit to the mall to perform more effective targeting.

While huge strides have been made in online analytics to extract a wealth of information from peoples’ online activities, corresponding work in the physical context — which we term as Physical Analytics — is relatively nascent and scattered. The goal of the proposed workshop is to bring together researchers and practitioners to have a conversation on Physical Analytics, with a view to coalescing a research agenda for the community. Our intention is keep the scope broad — spanning devices, algorithms, systems, applications, and policy — yet have the discourse be focused on physical analytics. So, for instance, within the scope of the workshop would be such topics as sensing, localization, wearable devices, cloud computing, data analytics, privacy, and more. However, rather than discussing the advances in these topics in isolation, the goal would be to focus on users, what we can learn about them, the algorithmic and systems issues involved in gleaning useful information, and how the resulting insights can be used for the benefit of users and businesses. Perspectives on how to balance the interests of users with those of businesses, the related economic issues, and what lessons the experience from online analytics holds for physical analytics, would also be welcome. However, much-studied problem areas such as sensor networks for monitoring the physical environment and big-data infrastructure, while certainly important, would not be in scope, unless a clear connection is made to user-centric physical analytics.

The workshop will include invited talks, a panel, and refereed papers, with as much time devoted to discussion as to the presentations themselves.

Keynote Speakers: Dina Katabi (Professor, MIT) and Anmol Madan (Co-Founder/CEO & Data Scientist, Ginger.io)

Invited speakers & panelists (pending travel confirmation, in some cases): Rajesh Balan (Associate Professor, Singapore Management University), Pravin Bhagwat (Founder & CTO, AirTight Networks), Tanzeem Choudhury, (Associate Professor, Cornell University), Jim Gemmell (CTO, Trov), and Robert Harle (Senior Lecturer, University of Cambridge)

Topics of interest Back to top

Note: This is not an exhaustive list.

         Wearable and mobile devices for physical analytics

         Continuous mobile vision and audio sensing

         Indoor localization applied to physical analytics

         RFID, NFC, cameras, and other infrastructure to support physical analytics

         Local and Cloud computing architecture

         Crowdsourcing for physical analytics

         Algorithms for analyzing and fusing physical signals, including location, video, and audio

         Personal data APIs and user privacy

         Interplay of physical analytics with online analytics

         Physical analytics in retail, healthcare, insurance, etc.

         Physical analytics to drive digital personal assistants and other applications

Program Co-Chairs and Committee Back to top

Program Co-Chairs

Deborah Estrin (Cornell Tech, USA)

Venkat Padmanabhan (Microsoft Research, India)

Contact: physicalanalytics2014-chairs@yahoogroups.com

Program Committee

Ramón Cáceres (AT&T Research, USA)

Romit Roy Choudhury (University of Illinois at Urbana-Champaign, USA)

Shyamnath Gollakota (University of Washington, USA)

Robert Harle (University of Cambridge, UK)

Fred Jiang (Intel Labs, China)

James Landay (Cornell Tech, USA)

Archan Misra (Singapore Management University, Singapore)

Matthai Philipose (Microsoft Research, USA)

Lili Qiu (University of Texas at Austin, USA)

Jacky Shen (Microsoft Research, China)

Junehwa Song (KAIST, Korea)

Roy Want (Google, USA)

Call for papers Back to top

We seek papers that report on work in progress, present an insightful survey of the state of the art, or lay out a compelling research agenda for the community, on all aspects of Physical Analytics, including devices, algorithms, systems, applications, and policy. The papers will be limited to 6 pages in length in the standard ACM format.

Important Dates

Submission deadline: March 24, 2014, 11:59 PM PDT (UTC-7)
Extended submission deadline: March 31, 2014, 11:59 PM PDT (UTC-7)
Notification of acceptance: April 16, 2014
Camera-ready deadline: April 27, 2014
Workshop date: June 16, 2014

Submission Instructions

Submission site URL: https://cmt.research.microsoft.com/PA2014/

All submissions must follow strictly the guidelines indicated below (adapted from the guidelines for the MobiSys 2014 conference).

  • Your submission must be in PDF. We will not accept the papers in any other format.
  • Your submission must use a 10pt font (or larger) and be correctly formatted for printing on Letter-sized (8.5" by 11") paper. Paper text blocks must follow ACM guidelines: double-column, with each column 9.25" by 3.33", 0.33" space between columns and single-spaced. If correctly formatted, this means that no page column will have more than 55 lines of text.
  • Submissions MUST be no more than six (6) pages. This length includes everything: figures, tables, references, appendices and so forth.
  • Provide an abstract of fewer than 1000 characters.
  • The first page of each paper should include the names and affiliations of the authors, i.e., the submissions should not be anonymous.
  • The paper must print clearly on standard black-and-white printers.
  • The PDF file size MUST be no more than 10 MB.