PhD Forum: HCI Applications

Location: South Ballroom

A Unified Approach to Computing and Visualizing Uncertainty in the Operating Room

Amber L. Simpson, Queen’s University

Information Visualization in Co-located Collaborative Environments

Petra Neumann, University of Calgary

Robust and Scalable Peer-To-Peer Video Distribution

Purvi Shah, University of Houston

PhD Forum: Algorithms for Parallel and Distributed Computing

Location: Salon III

Massively Parallel Sequence Alignment for the Bioinformatics Domain

Shannon I. Steinfadt, Kent State University

On Node Isolation in a Directional Wireless Sensor Network

Unoma N. Okorafor, Texas A & M University

New Investigators: Knowledge Discovery & Data Mining

Location: Salon VI and VII

Maximizing the Extent of Spread in a Dynamic Network

Habiba Habiba (University of Illinois at Chicago), Tanya Y. Berger-Wolf (University of Illinois at Chicago)

We focus on the problem of identifying a set of individuals to initiate the spreading process in a dynamic population of interacting individuals so that the eventual extent of the spread is maximized. We extend a spreading approach on static networks to explicitly dynamic networks. Maximizing the extent of spread in a dynamic network is NP-hard. We present and evaluate on real data an approximation algorithm for solving this problem.

New Investigators: Mathematical Programming

Location: Salon VIII

On the Upper Bound of Sensor Network Lifetime – A Flow Optimization Approach

Tuli Nivas (University of Texas at Arlington), Gergely V. Zaruba (University of Texas at Arlington)

The growth in technology has made sensor networks(WSN)a reality. In such networks, energy efficiency of nodes and protocols is the most paramount issue. This paper deals with the problem of finding upper bounds for lifetimes in sensor networks. We use a custom simulated annealing optimization method to derive the upper bound for our framework. We also demonstrate how to use our lifetime bounds to evaluate sensor network protocols.

A New Application of Linear Algebra in Latent Semantic Indexing

Jane E. Tougas (Dalhousie University)

Latent semantic indexing is an information retrieval method that represents a document collection by a term-document matrix. This numerical representation is then reduced in size by factorizing the matrix. Recomputing this factorization whenever a document collection changes is expensive. This paper discusses alternatives to the recomputation, focusing on a combination of two methods, called folding-up, and using a linear algebra error measure to determine when to switch between these methods.

New Investigators: Theory & Systems

Location: Narcissus and Orange Blossom

Sherlock: Tracking Down Propagating Problems in Distributed Systems

Soila Pertet (Carnegie Mellon University), Rajeev Gandhi (Carnegie Mellon University), Priya Narasimhan (Carnegie Mellon University)

Distributed systems contain multiple components that can interact in sometimes unforeseen and complicated ways; increasing the likelihood of propagating problems that might result in widespread disruption. Mitigating the impact of these problems involves “fingerpointing”, i.e., determining the root-cause of the failure, coupled with failure-recovery strategies. We present Sherlock, an approach for fingerpointing propagating problems in distributed systems, and discuss our initial work at fingerpointing performance problems in replicated systems.

It’s Not Magic: I Can Prove It

Anya Tafliovich (University of Toronto)

Our work presents a new approach to developing, analysing, and proving correctness of programs intended for execution on a quantum computer. We provide tools to write quantum as well as classical specifications, develop quantum and classical solutions for them, and analyse various properties of quantum specifications and quantum programs, such as implementability, time and space complexity, and probabilistic error analysis uniformly, all in the same framework.