Research and Review Methodology: The Way Our Community Continues to Thrive

• I Aver: Providing Declarative Experiment Specifications Facilitates the Evaluation of Computer Systems Research
  
  Providing declarative statements that describe the outcome of an experiment can significantly improve the task of validating its results.
  Ivo Jimenez (UC Santa Cruz)

  Carlos Maltzahn (UC Santa Cruz)

  Jay Lofstead (Sandia National Laboratories)

  Adam Moody (Lawrence Livermore National Laboratory)

  Kathryn Mohror (Lawrence Livermore National Laboratory)

  Remzi Arpaci-Dusseau (University of Wisconsin-Madison)

  Andrea Arpaci-Dusseau (University of Wisconsin-Madison)
• It is More Blessed to Give than to Receive - Open Software Tools Enable Open Innovation
  
  Sharing software tools enables open innovation, brings faster upgrades and frees up resources, but demands investments in the open community.
  Per Runeson (Lund University)

  Hussan Munir (Lund University)

  Krzysztof Wnuk (Blekinge Institute of Technology)
• The Full Mont(olog)y: Using Ontologies to Optimize the CS Literature Review
  
  We need a comprehensive and maintainable ontology of CS research to drive scientific progress despite massive knowledge availability.
  Kathryn Dahlgren (UC Santa Cruz)

Power, Energy, and Thermal: How Do We Survive in the Post-Dennard Scaling Era

• Squeezing Energy Savings Out Of Similar Data and Computation in GPGPUs
  
  Significant energy savings (with <5.5% output quality loss) can be achieved by merging computation and storage of similar data.
  Daniel Wong (UC Riverside)
• Reducing Mobile Throttling from Temperature by Offloading Apps (★Video)
  
  Offload intensive apps from phones to avoid high temperatures that cause throttling and reduce global performance.
  William Agnew (Georgia Institute of Technology)

  Hyunjong Lee (Georgia Institute of Technology)

  Minsung Jang (AT&T Research Lab)

  Ada Gavrilovska (Georgia Institute of Technology)

  Karsten Schwan (Georgia Institute of Technology)

Knowledge Discovery and Learning: Drawing Insights from "Big Data"

• Machine Learning Literature using Textual Features
  
  Why are some novels considered to be literature? A predictive model of literary judgments shows that textual features are important.
  Andreas van Cranenburgh (Huygens ING & Institute for Logic, Language and Computation)
• Data Integration Beyond Alignments Between Two Sources
  
  Reconciling identity vs. distinctness of items from different sources can be NP-hard, but approximations enable flexible data integration
  Gerard de Melo (Tsinghua University)
• Is Small More Interesting? Examining Countries' GeoNames Places linked to Wikipedia
  
  Places in small countries are three times as likely to be linked to Wikipedia with 41% of coverage compared to 13% for larger countries.
  Dirk Ahlers (NTNU - Norwegian University of Science and Technology)
• Do Take it Personal: It's Not What You Say, It's Who (and Where) You Are!
  
  Priority of reported issues is a poor predictor of future code change. Suspicions empirically confirmed: submitter metadata most important.
  Markus Borg (SICS Swedish ICT AB)

  Petter Gulin (Lund University)

  Linus Olofsson (Lund University)

Security and Privacy: Issues When Client, Cloud, and Crowd are Closely Integrated

• Improving Android security by widening the role of Mandatory Access Control
  
  The Android security ecosystem can greatly benefit from a deeper integration of SELinux Mandatory Access Control into system components.
  Enrico Bacis (Universitá degli Studi di Bergamo)

  Simone Mutti (Universitá degli Studi di Bergamo)

  Marco Rosa (Universitá degli Studi di Bergamo)

  Stefano Paraboschi (Universitá degli Studi di Bergamo)
• Investigating User Profiling and Privacy Leaks in Mobile Ad Networks
  
  User-based targeting in mobile ads is scarce & we explore if attackers can track a user by spoofing the user's identifiers in ad requests.
  Eeva Terkki (University of Helsinki)

  Ashwin Rao (University of Helsinki)

  Sasu Tarkoma (University of Helsinki)
• On Information Flow Control and Audit for Demonstrable Compliance in the Cloud
  
  Complex policy can be built upon simple primitives enforced on every data flow. Enforcement data can be captured to demonstrate compliance.
  Thomas Pasquier (University of Cambridge)

  Jatinder Singh (University of Cambridge)

  Jean Bacon (University of Cambridge)
• A Network-Aware Access Control System for Providing Good Enough Security to Low Value Information
  
  Network-aware access control systems can provide good-enough security for sharing information with low value over intermittent network.
  Reena Singh (Indian Institute of Technology, Mandi)

  Timothy A. Gonsalves (indian institute of technology, Mandi)

Computer Networks: SDN, 5G, and Beyond

• How real is the threat of instability events predicted by Adversarial Queueing Theory?
  
  Theoretical results suggests that innocent traffic patterns can overwhelm networks. Yet, for finite buffers these patterns appear harmless.
  Daniel S. Berger (University of Kaiserslautern)

  Martin Karsten (University of Waterloo)

  Jens Schmitt (University of Kaiserslautern)
• Towards Secure Decoy Routing by Using SDN
  
  Against existing censorship circumvention protocols, we propose SDN-based Decoy Routing that is faster and resilient to traffic analysis.
  Chaitanya Kumar (IIIT-Delhi)

  Vinayak Naik (IIIT-Delhi)
• Solving Signaling Storm through Long-Term Evolution Refactoring
  
  Refactoring LTE networks opens a gold mine of research problems and provides insights for laying the cornerstone for modular 5G networks.
  Matteo Pozza (University of Helsinki, University of Padova)

  Heikki Lindholm (University of Helsinki)

  Hannu Flinck (Nokia Networks)

  Claudio Palazzi (University of Padova)

  Ashwin Rao (University of Helsinki)

  Sasu Tarkoma (University of Helsinki)

Data and Data Structures: Making Better Software by Understanding Data Better

• A Framework for Testing Properties of Discrete Distributions: Monotonicity, Independence, and More
  
  Wanna optimally test if your distribution has a property? Assume it does, learn the distribution, and test the hypothesis. Use χ² distance!
  Jayadev Acharya (EECS, MIT)

  Constantinos Daskalakis (EECS, MIT)

  Gautam Kamath (EECS, MIT)
• QP tries are smaller and faster than crit-bit trees
  
  Typical QP trie depth is 0.35-0.40 space is 0.5-0.6 time is 0.6-0.8 of equivalent crit-bit tree.
  Tony Finch (University of Cambridge)
• Programming with Enumerators: Constructive Approach to Challenge the Status Quo in Automated Data Structure Generation
  
  We define enumerators that assign numbers to arbitrary values. Their composition efficiently enumerates big sets of complex data structures.
  Ivan Kuraj (CSAIL, MIT)