|
 |
USER Summer 2010 Internships
|
The User Systems & Experience Research (USER) group at IBM Research - Almaden in San Jose, California is accepting applications for interns for the summer of 2010. The USER group focuses on understanding and improving how people interact with technology. Our goal is to improve the ease-of-use of existing products and explore new paradigms in using computers. Our cross-disciplinary team is comprised of human computer interaction (HCI) and human factors researchers, software engineers, computer security experts, cognitive and social computing scientists, user interface designers, and user study analysts.
Our current research focuses include supporting user activities, improving collaboration within the enterprise, enhancing browser-based user experiences, interaction with multiple computing devices, and user experiences for mobile devices. We seek interns with backgrounds in computer science, cognitive science, sociology, and design. Applicants must be enrolled in a Ph.D. program.
To apply for an internship, send an email to csjobs_user at almaden.ibm.com:
- with a subject-line "USER summer intern 2010 application"
- attaching your CV, and
- including the following information in the body text of the email:
- A few keywords describing your current research and research area (e.g., CSCW, info viz, end-user-programming, experimental psychology)
- A few keywords describing your HCI research skills (e.g., strong programmer, quantitative data analysis, experimental design, visual design)
- The name and email address of your academic advisor
Applications sent to csjobs_user at almaden.ibm.com will be considered for all USER research projects. If you know which project you are interested in, you are encouraged to send an email with the above information to the contact person for that individual project (listed below) -- your application will still be considered for all USER projects. Please do not send your application directly to multiple project teams.
IBM is committed to creating a diverse environment and is proud to be an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, gender, gender identity or expression, sexual orientation, national origin, genetics, disability, age, or veteran status.
|
|
|
|
Stroke gestures are an increasingly important interaction medium for mobile and other touch surface-based computing devices. Research at Almaden in this area includes ShapeWriter, an advanced text entry system, and modeling of stroke gesture (the CLC model of stroke gestures and the steering law). Depending on the candidate's interests and skills, this summer project should make contributions in one of the following directions: 1. Mathematical / statistical / information theoretic understanding of gesture stroke systems such as ShapeWriter as a function of information capacity (e.g. the number of words and commands), 2. Advancing stroke gesture interface speed and accuracy performance, 3. Empirical and analytical studies of stroke gesture systems. The candidates should have skills and interest in at least two of the following: 1. Mathematical and statistical analysis and modeling, 2. Programming and prototyping, 3. Human performance literature and empirical methods.
Team: Shumin Zhai, Huahai Yang
Contact: Shumin Zhai (zhai at almaden.ibm.com)
|
|
The CoTester research team is exploring technology to make Web testing lightweight and easy for testers and end users. CoTester builds upon the CoScripter approach and allows testers to create test scripts that are represented in an easy-to-understand scripting language instead of a complex programming language. The use of this language, dubbed ClearScript, enables a wider variety of testers to create and understand scripts without the need for programming skills. In addition, we have designed and implemented a machine learning algorithm to automatically identify subroutines in test scripts. Such an algorithm can group together a sequence of low-level actions within a test into a conceptual unit representing a higher-level action.
In summer 2010, we would like to explore whether our algorithm can improve test script maintenance and lower the barrier of website testing. Specifically, we expect that the intern will:
- Build an interface that can automatically incorporate the changes in test scripts as a result of webpage changes.
- Experiment whether and how subroutines can help test script maintenance.
- Possibly improve our algorithm and apply other machine learning techniques.
- Conduct a user study.
An ideal candidate should have the following skills: knowledge of Java and Javascript programming; solid background in algorithms and machine learning. HCI background is a plus.
Team: Allen Cypher, Clemens Drews, Anca Ivan, Tessa Lau, Jalal Mahmud, Jeffrey Nichols
Contact: Jalal Mahmud (jumahmud at us.ibm.com)
|
|
As users increasingly employ highly capable mobile devices such as the iPhone, Palm Pre, and Android devices, we have an opportunity to rethink the user experiences, particularly how users interact with mobile devices, both in isolation and in the context of their other computing devices (e.g., desktops, laptops). We are currently exploring:
- How to redesign both mobile and desktop/laptop user experiences to explicitly recognize and reflect the use of multiple, heterogeneous devices for computational work. For example, how could/should email clients change when users employ both a mobile device and a desktop or laptop to check their mail?
- How to support computational activities that span devices. We are particularly interested in how different patterns of behavior on different devices (e.g., short, intermittent use on mobile devices vs. sustained use of desktops) can inform both the design of new user experiences and the requirements for an underlying infrastructure.
We are looking for an intern to explore and prototype new user experiences that span multiple computing devices.
Contact: Jeff Pierce (jspierce at us.ibm.com)
|
|
The current set of user experiences and user interfaces for mobile devices are just scratching the surface; we're interested in exploring new opportunities to extend the types and ways that users employ their mobile devices. We are looking for an intern who is interesting in pushing the boundaries of interaction for mobile devices to create compelling next-generation experiences.
Contact: Jeff Pierce (jspierce at us.ibm.com)
|
|
Node-link graph visualization appears in many application domains, including 2D vector based road maps, social networking visualization, and visualizing entity relationships in databases, and more. A key problem in large-scale node-link graph visualization is how to avoid visual clutters and reduce information overload. We take a user centered generalization approach to solving these problems. Generalization techniques may remove, merge, alter, and modify some parts of the graph in order to emphasize important information while still represent the graph in a faithful and recognizable way. One of the key focuses of this project is to design interactive generalization techniques that tailor to specific user tasks.
Contact: Huahai Yang (hyang at us.ibm.com)
|
|
Highlight allows end users to create mobile applications from existing web sites and deploy those applications to mobile devices. Highlight's design, which incorporates a fully-functional modern web browser engine as part of its server-side infrastructure, allows existing sites that contain substantial client-side JavaScript or Ajax technology to be modified for use on mobile devices. Our end-user authoring environment makes use of programming-by-demonstration techniques to allow creation of mobile applications easily and quickly.
Team: Jeffrey Nichols, John Barton, Tessa Lau
Contact: Jeffrey Nichols (jwnichols at us.ibm.com)
|
|
This project explores effective ways of improving English vocabulary of students, particularly children with English-as-second-language. The ideal intern student should have in-depth knowledge or interest in memory and education psychology, as well as skills in interface design and research. Mobile application development skills are a plus but not required.
Contact: Shumin Zhai (zha at almaden.ibm.com)
|
|
We are combining end-user programming concepts from the CoScripter project with the Firebug website debugging tool to create a new, multi-user web user interface development system. We want this system to encourage development of Web UIs that work well on mobile and desktops because they adapt at the task-level rather than layout level.
Contact: John Barton (bartonjj at us.ibm.com)
|
|
PreMa aims to empower end users with tools to create new data visualization using existing visualization as raw materials. Users are alleviated from the burden of digging into complex data repository and do not have to already be familiar with the intricacies of visual analytics. PreMa takes advantage of semantic search capabilities currently available, allowing users to select visualization artifacts from search results and drag and drop them together to produce new visualization. Under the hood, PreMa uses an algebraic data presentation language engine to construct meaningful chart compositions. PreMa uses rich user interface feedback to guide the user making decisions on the construction of chart compositions.
Contact: Huahai Yang (hyang at us.ibm.com)
|
|
We are exploring the use of social and collaboration software within the enterprise, including email, social networks, microblogging, and contact management.
- Bridging the gap between email and social software: Exploring topics such as social features (like tagging and profiles) in email, attention management, and bringing email into social applications.
- Collaboration context: Using social and other cues to bring a context around a current piece of work, making it easier to find related information when you're working on something.
- Awareness: Using microblogging and automatic updates to improve general awareness of what your colleagues are working on.
- Unified contact management: Exploring how people can keep their social network intact between applications and how people work with groups.
- Cloud-based collaboration storage: Using a Cassandra-based cloud to provide a scalable backend for email, feeds, and other collaborations.
Team: Stephen Dill, Eric Wilcox, Jan Pieper, Jerre Schoudt, Julia Grace, Julian Cerruti, Sandra Yuen
Contact: Stephen Dill (dill at us.ibm.com)
|
|
Computer system administrators are the unsung heroes of the information age, working behind the scenes to configure, maintain, and troubleshoot the computer infrastructure that underlies much of modern life. However, little can be found in the literature about the practices and problems of these highly specialized computer users. Through a series of field studies in large corporate data centers, observing organizations, work practices, tools, and problem-solving strategies of system administrators, we observed that modern IT systems are extremely complex and a single individual cannot manage the complexity of such systems. Though organizations, methods, and tools are often put in place to reduce complexity in IT management, we found that imposing top-down structure in organizations, methods, and tools can only trail innovation in IT systems. We argue that real productivity gains can only be obtained when individuals can invent their own methods and tools for managing IT complexity and when these methods and tools are adopted by local teams and by larger organizations. Based on these findings, we are exploring new tools and features to improve administrators' ability to collaboratively manage complex systems.
Team: Eser Kandogan, Eben Haber, John Bailey, Paul Maglio, Tara Matthews, Andreas Dieberger
Contact: Tara Matthews (tlmatthe at us.ibm.com)
|
|
This project will explore the use of personalized and collaborative task models to customize Web pages and present relevant concepts. Prior research has demonstrated that such models can substantially improve the Web transaction experience, such as shopping, for visually-challenged and mobile handheld users. We would like to further explore how such models can be used to facilitate general Web browsing task, such as reading newspapers or booking travel, for regular users. The browsing experience can be enhanced based on the recommendation given by the task model and models can be parameterized based on labeling of Web page elements, users preference, past browsing history, and context. We will explore both theoretical framework of such models and their usability issues.
The candidate should have strong data mining and machine learning experience. Prototyping and programming (Java, Java Script) is also required.
Contact: Jalal Mahmud (jumahmud at us.ibm.com)
|
|
Teams have many collaboration tools to help them organize project materials in one place, where all members can view them. Despite these tools, teams still revert to email for communication and file sharing. Email is always on, whereas a collaboration tool feels like "yet another tool." To preserve the benefits of collaboration tools and the convenience of email for communication, we have designed Topika, a web service that allows a user to collect and share email about any topic, and have it automatically organized within their collaboration tool of choice. For the user, communicating with teammates and routing those threads into collaboration tools is as simple as sending and receiving emails. To facilitate this, Topika leverages knowledge of the user's collaboration practices and machine learning.
The Topika project is seeking students with strong programming skills and machine learning expertise. An interest in CSCW is a plus.
Team: Tom Moran, Jalal Mahmud, Tessa Lau, Tara Matthews, Barton Smith, Steve Whittaker
Contact: Jalal Mahmud (jmahmud at us.ibm.com)
|
|
Highly capable mobile devices such as the iPhone are changing how people engage with computing. We've been studying our internal population of iPhone users to better understand mobile usage practices. While we've learned a number of valuable lessons, there are lots of opportunities to expand our efforts to explore usage in a wider variety of contexts, for a wider variety of tasks, and by more varied user populations.
We are looking for an intern to help us study and better understand mobile device use.
Contact: Jeff Pierce (jspierce at us.ibm.com)
|
|
Many business applications would be better served with customized results, such as personalized search results during search, selective sharing of information with people, and people, tools and content recommended for individual users. To better serve individual needs, we are developing a unified framework that dynamically builds a profile that reflects an individual's or a team's communication and collaboration practices. The profile includes individual/team interests, skills, preferred tools, membership in shared online spaces, and more. We are working on algorithms and tools that can interactively and automatically build such a individual or team profile from various sources (including socio-collaborative tools) and automatically update the profile based on user feedback.
Team: Hongxia Jin, Stephen Whittaker, Qihua Wang
Contact: Hongxia Jin (jin at us.ibm.com)
|
|
Most work in social network analysis has focused on the nature and structure of people's networks in their everyday social lives. Instead, we are looking at people's work based social networks. We hope to address the problems that workers face when managing, communicating, and collaborating with their work based social network. For example, information-dense email thread are often lost in overloaded email inboxes and not shared with all relevant subsets of one's work social network. To explore solutions to such problems, we will focus on the following research questions: What are the meaningful subsets of an individual's work social network and what functions do these subsets serve? Specifically, as a subset, what is their: Makeup (i.e., how many / who are the members)?; Purpose for existing?; Relationship to each other?; Manifestations in various social networking, communication, and collaborative tools (i.e., do they email each other, have a shared wiki, etc.)? We will apply our findings to develop algorithms that extract these subsets automatically (e.g. by email analysis or by mining other social network tools) and to improve workers' experience collaborating with subsets of their social network.
Team: Steve Whittaker, Tara Matthews, Julian Cerruti, Steve Dill, Eric Wilcox
Contact: Steve Whittaker (sjwhitta at us.ibm.com)
|
|
Current online collaboration tools offer limited flexibility when organizing work, which makes it difficult to use these tools for the multiple phases of a complex project, such as brainstorming, deciding upon concrete goals and tasks, working closely with teammates on those tasks, and reporting final outcomes. To support multiple phases of collaborative projects, we have designed Workboards, a shared, online tool providing incredibly flexible spatial layouts of content, quick and easy editing, support for synchronous and asynchronous collaboration, and the ability to repurpose the large quantities of content generated during a project.
The Workboards project is seeking students with strong programming skills. An interest in CSCW and/or information visualization are a plus.
Team: Tom Moran, Eser Kandogan, Eric Wilcox, Sandra Yuen, Pablo Pedemonte
Contact: Eser Kandogan (eser at us.ibm.com)
|
|
|
