Mentor: Deb Agarwal (Lawrence Berkeley National Laboratory)

A Flexible, Scalable Framework for Integrating Heterogeneous Sequence Data in Comparative Genomics

Presenter: Allison A. Regier (University of Notre Dame)

The assembly step is critical for using genome sequence data, but it may introduce errors and/or hide ambiguities. Previously, the underlying sequence was discarded because of its high computational demands. I am developing a framework to efficiently access underlying sequence data during downstream analysis. The framework leverages new developments in distributed computing that make common data access patterns efficient. We will use this information to improve comparative genome analysis.

———-
Multivariate Time Series Analysis of Clinical and Physiological Data

Presenter: Patricia Ordóñez Rozo (University of Maryland, Baltimore County)

We aim to create a multivariate temporal representation of electronic medical data which automates the personalization of baselines and thresholds based on a patient’s history. Visualizations based on the representation emphasize the rate of change in variables and assist providers in analyzing the data from a multivariate perspective. A novel similarity metric for this representation will be the cornerstone to the development of a search engine for large medical databases.

———-

Multi-Agent Fault Tolerance Inspired by a Computational Analysis of Cancer

Presenter: Megan Olsen (University of Massachusetts, Amherst)

In cancer biology, it is known that cancer cells can disappear without therapy, but not how. We propose that cells communicate such that primarily malfunctioning cells (tumors) die. We also propose that this same communication can be used as inspiration for a fault-tolerance mechanism for multi-agent and distributed systems to remove faulty agents using only local information. I examine the communication protocols necessary for removing these faults in both systems.

Mentor: Patty Lopez (Intel)

Predictor Virtualization: Teaching Old Caches New Tricks

Presenter: Ioana M. Burcea (University of Toronto)

We present Predictor Virtualization (PV), a technique that takes advantage of the existing memory hierarchy (i.e., processor caches and main memory) to emulate large prediction tables for hardware optimizations. PV increases the utility of traditional caches: in addition to being accelerators for slow off-chip memories, the on-chip memory hierarchy becomes leverage for effective predictor-based hardware optimizations.

———-

Coordinated System Level Resource Management for Heterogeneous Many-core Platforms

Presenter: Vishakha Gupta (Georgia Institute of Technology)

A challenge posed by future architectures is the efficient exploitation of their many and sometimes heterogeneous cores. This is exacerbated by multiple facilities for data movement and sharing across cores on such platforms. Our work aims to enable high performance program execution and efficient resource utilization in such platforms. Hence, we propose to virtualize platforms to allow for flexibility in targeting functionality, schedule VMs efficiently and create underlying system-level technologies.

———-

Throughput-Driven Optimizations for Programming Multi-Core Platforms

Presenter: Rebecca Collins (Columbia University)

Multi-core architectures are ubiquitous today, and there is a need for high level programming tools that capture an application’s parallel substructure without placing too great a burden on the programmer. My research includes two domain-specific tools that raise the level of programming abstraction while enhancing system throughput for data-driven applications.

Mentor: Andrea Danyluk (Williams College)

Hybrid Methods for Generating and Evaluating Style-Specific Accompaniment

Presenter: Ching-Hua Chuan (University of North Florida)

Creating distinctive harmonizations in an identifiable style may be one of the most difficult tasks for amateur song writers. To model and assist in this creative process, we present a hybrid computer system combining knowledge of musical theory and statistical learning. The system is capable of learning a style from only a few examples to create new harmonizations for melodies with proper harmonic resolutions. Experiments were conducted using pop-rock songs.

———-

Supporting Collaborative Learning in Narrative-Centered Learning Environments

Presenter: Jennifer L. Robison (North Carolina State University)

Student collaboration in traditional classroom settings has been shown to have multifaceted benefits for students. However, creating effective cooperative learning environments is challenging. The intelligent tutoring systems community has begun to investigate the utility of computer-supported collaborative learning as a means for addressing these challenges. This paper discuss the advantages of intelligent tutoring systems, with a focus on narrative-centered learning environments, for supporting effective student collaboration.

———-

The Effect of Age, Gender, and Previous Gaming Experience on Customization Activities Within Games

Presenter: Mona Erfani Joorabchi (Simon Fraser University)

Understanding players’ game playing behavior is a growing area of research currently being explored by many game companies. In this paper, we report on a study we conducted to understand the influence of age, gender, and previous gaming experience on customization activities of game players in the younger age group within games. We note that player behavior can be used to improve game design and may lead to new player models.

———-

Designing Technology to Support Nurses’ Information Flow

Presenter: Charlotte Tang (University of Calgary)

This research comprised of longitudinal field studies that contributed better understanding of nurses’ information flow dynamics, and impacts of new technologies. It contributed the InfoFlow Framework for evaluating new technologies and for generating new technology designs, as well as a paper-and-digital integrated charting prototype. A focus group conducted to evaluate this prototype indicated promising potential for bridging nurses’ use of informal paper artifacts and the organizational digital medical records.

Mentor: Nancy Cam-Winget (Cisco)

An Application of Clustering Techniques to Urban Studies

Presenter: Zahra Ferdowsi (DePaul University)

This paper demonstrates the usefulness of data mining techniques in uncovering socioeconomic characteristics of modern cities and in understanding social processes affecting urban life. K-means clustering techniques and decision tree analysis were used to create a typology of urban neighborhoods. The data-driven typology was able to capture the spatial relationship between areas with similar socioeconomic profiles, showing that people with common social status tend to aggregate in adjacent areas.

———-

Retrofitting Authorization In Legacy Programs

Presenter: Divya Muthukumaran (The Pennsylvania State University)

Programs need to enforce security policies on behalf of users (e.g., secrecy, integrity and runtime access check), and need to be retrofit with mediation statements to support this. In recent work, we have found that placing mediation points in programs is equivalent to computing a graph-cut, so we propose to apply this technology along with placement policies to automate the task of retrofitting programs for authorization.

———-

You Are What You Search: Query-Based Data Mining for the Web

Presenter: Barbara Poblete (Yahoo! Research)

In this work we take a look into the behavior of Web users, in particular into the queries that they submit to search engines. Queries provide a unique insight into users’ needs and goals while surfing the Web. This information is provided by users implicitly and is key for improving Web sites, search engine raking results, as well as the organization of the Web in general.

———-

A New Relational Naive Bayes Net Classifier

Presenters: Bahareh Bina (Simon Fraser University)

Relational classification is the problem of predicting the class attribute of a target entity given its attributes and related entities’ attributes. In this paper we propose a new relational naive Bayes classifiers. Previous work assume links independent and put the same weight for attributes of the target entity and related entities. We modify this assumption and propose a new method that places more weight on the target features than links’ features.

Mentor: Andrea Danylnk (Williams College)

Usable Security in Practice: Collaborative Management of Electronic & Physical Personal Information

Presenter: Laurian C. Vega (Virginia Polytecnial Institute)

A need exists in HCI to study how issues of trust and privacy can and do affect the ad hoc negotiation of security rules and how they are managed in actual practice. For my dissertation I present data from pilot interviews and observations to examine the physical and electronic security practices of childcares and medical offices. I propose to study the policy breakdowns that affect the security of personal information.

———-

In the Beginning: The Early Lives of Users in Online Communities

Presenter: Katherine Panciera (University of Minnesota)

The most crucial part of a user’s life is the time between when they first view a system and twenty-four hours after their first participation in that system. From that point on, odds are that their participation will decline. When designing online communities, we need to understand this more deeply. My research also proposes a mixed methods study of early life user behavior in Cyclopath, a geowiki for bicyclists.

———-

Building Professional Identity as Computer Science Teachers: Supporting Secondary Computer Science Teachers through Reflection and Community Building

Presenter: Lijun Ni (Georgia Institute of Technology)

Currently, we are facing big challenges of preparing and supporting K-12 CS teachers. In addition to increasing the number of CS teachers, there is a great need of supporting those teachers to grow and retain as committed, quality teachers. This thesis work focuses on exploring ways of supporting CS teachers through understanding their teacher identity and exploring ways of supporting identity development through a professional development program for CS teachers.

Mentor: Elizabeth Mynatt (Georgia Institute of Technology)

Word Alignment with Parallel Phrases

Presenter: Maria Holmqvist (Linköping University)

I present a method for word alignment that uses parallel phrases from manually word aligned sentence pairs to align words in new texts. Experiments on an English-Swedish parallel corpus showed that the phrase-based method produced word alignments with high precision. Phrase-based alignment was compared to statistical word alignment and it was found that a combination of phrase-based and statistical word alignments outperformed pure statistical alignment in terms of Alignment Error Rate (AER).

———-

Modeling and Testing of Service-Oriented Applications

Presenter: Ewa Musial (University at Albany, State University of New York)

In this paper we present a technique testing not only services but also their interactions with execution context. The technique utilizes a dependence model that depicts relationships between services, external resources, and other application components. Based on this model, we first associate the specifications with the model to cover the requirements. Then we use node and dependence edge coverage criteria to create additional test cases to achieve adequate testing.

———-

Assuring Network Service with Bandwidth and Integrity Based Fairness

Presenter: Fariba Khan (University of Illinois at Urbana-Champaign)

DDoS attackers take control of large bandwidth and services that otherwise would have been lightly used by genuine users. IBQ and ASV take the power back from them by taking two different approaches. In ASV clients use otherwise underutilized bandwidth. ASV can be implemented, without the cooperation of the core routers, by slight modification of the client and server applications.IBQ introduces a tiered inegrity system and prioritizes high integrity packets.

Mentor: Ellen Walker (Hiuram College)

A Mobile-Cloud Collaborative Approach for Context-Aware Blind Navigation

Presenter: Pelin Angin (Purdue University)

Context-awareness is a critical aspect of safe navigation, especially for the blind in unfamiliar environments. Existing devices for context-aware navigation fall short due to their dependence on infrastructure requirements and limited access to resources of contextual clues. In this paper, we propose a mobile-cloud collaborative approach for context-aware navigation by exploiting resources of Cloud Computing providers and location-specific resources on the Internet. We propose an extensible architecture allowing for wide usability.

Presenters: Connie Smallwood (CA Technologies), Matt Millunchick (Facebook), Dawn Carter (Amazon), and Shana Venson (Lockheed Martin)

Whether you are looking for your first job or thinking about a career change, come to this session to learn how to find a job. This panel discussion will cover all the activities that go into getting to a job interview, including: resume and cover letter writing tips; how to engage your network in your job search; and what to do when you attend a job fair.

Panelists: Karen Weiss (Intuit), Jennifer Hall (Intuit), Krista Claude (Thomson Reuters), Sinead Strain (Goldman Sachs), Melody Browne (Deutsche Bank), and Hanna Derry (BlackRock)

It is estimated that 1 in 4 workers regardless of sector will be a technologist. Explore career choices as this panel presents innovative technology solutions within industries. What is quantitative analysis? How are traders supported at Goldman Sachs? What is an information services company like Thomson Reuters developing to remain competitive? How is Intuit developing innovative small business software solutions? Each offers interesting and challenging technical careers.

Presenter: Anna Krasnyanskaya (Symantec), Kahleen Naughton (HP), and Divya Kolar (Intel)

This panel discussion will cover your interactions with your potential employer including: how to master the technical interview; writing the thank you note; making sure your references are good ones; and how to negotiate the job offer. Get the recruiters’ and interviewers’ perspectives on what they are looking for from technical candidates. Bring your questions along for the Q&A session.

Panelists: Pamela Arya (ABI Advisory Board and A-T Solutions), Lori Hecker (Raytheon), Kathleen A. Hutson (National Security Agency), Bonnie Kean (Northup Grumman), Tamara Yu (MIT Lincoln Laboratory), Rear Admiral Gretchen A. Herbert (US Navy), Rebecca Wilems (Lockheed Martin), Rear Admiral Janice M. Hamby (C4 Systems, Joint Chiefs of Staff), and Sue Dragich (Lockheed Martin)

b. Some of the most advanced technology is being developed in defense and cyber security. Interested in cybercrime technology to safeguard systems vulnerable to viruses, hacking and identity theft? Have you thought of becoming a Computer Systems Analyst or Information Security Analyst? Are you interested in what it takes to obtain a security clearance? If so, please join this workshop to discuss technology innovation, career opportunities and current challenges within US defense and cyber security.