Cynthia Taylor

The RUBI project: a progress report

The goal of the RUBI project is to accelerate progress in the development of social robots by addressing the problem at multiple levels, including the development of a scientific agenda, research methods, formal approaches, software, and hardware. The project is based on the idea that progress will go hand-in-hand with the emergence of a new scientific discipline that focuses on understanding the organization of adaptive behavior in real-time within the environments in which organisms operate. As such, the RUBI project emphasizes the process of design by immersion, i.e., embedding scientists, engineers and robots in everyday life environments so as to have these environments shape the hardware, software, and scientific questions as early as possible in the development process. The focus of the project so far has been on social robots that interact with 18 to 24 month old toddlers as part of their daily activities at the Early Childhood Education Center at the University of California, San Diego. In this document we present an overall assessment of the lessons and progress through year two of the project.

Computer science concept inventories: past and future

Concept Inventories (CIs) are assessments designed to measure student learning of core concepts. CIs have become well known for their major impact on pedagogical techniques in other sciences, especially physics. Presently, there are no widely used, validated CIs for computer science. However, considerable groundwork has been performed in the form of identifying core concepts, analyzing student misconceptions, and developing CI assessment questions. Although much of the work has been focused on CS1 and a CI has been developed for digital logic, some preliminary work on CIs is underway for other courses. This literature review examines CI work in other STEM disciplines, discusses the preliminary development of CIs in computer science, and outlines related research in computer science education that contributes to CI development.

Improving video performance in VNC under high latency conditions

Virtual Network Computing, or VNC, is a popular thin client application used to access files and applications on remote computers. However, VNC can suffer from significant losses in throughput when there is high latency between the client and server. These losses become especially apparent in the case of video, where updates are both large and frequent. In this work, we present a Message Accelerator proxy for VNC. This simple but highly effective solution mitigates high latency network effects for video performance while maintaining the advantages of a client-pull system. By operating near/on the server, it can send updates to the client at a rate corresponding to proxyserver interactions which are faster than client-server interactions. When testing using video, our Message Accelerator design results in frame rates an order of magnitude higher than plain VNC when running under high latency conditions.

Towards a Proximal resource-based architecture to support augmented reality applications

We are developing a new enhanced cloud-based computing architecture, called the Proximal Workspace architecture to allow access and interaction between lightweight devices, e.g., video glasses, earphones, wrist displays, body sensors, etc., and applications that represent a new generation of computationand-data-intensive programs in areas such as augmented reality. While lightweight devices offer an easy way for these applications to collect user data and offer feedback, the applications cannot be run natively and completely on these devices because of high resource demands. Making these applications available via a cloud, while promoting ubiquitous access and providing the necessary resources to execute the applications, induces large delays due to network latency. To solve these problems, we are developing a new system architecture based on supporting workspaces which provide nearby computing power to the users devices and thus mediate between them and the clouds computing resources. Specifically, a workspace provides a set of middleware utilities designed to exploit local resources, and provide specific functions such as rendering of graphics, pre-fetching of data, and combining data from different servers. More generally, the workspace is designed to run any subset of activities that cannot be run on a users device due to computation speed or storage size, and cannot be run on a cloud server due to network latency. Ultimately, the goal is to produce a set of middleware utilities that when run in the workspace with highly-interactive, computation/data intensive applications will result in better user-perceived performance. We are exploring this system architecture constructively, by adapting applications that benefit from this architecture and discovering how best to suit their needs. We have already adapted VNC (Virtual Network Computing, which allows local interaction with remote computations) to run under a similar architecture, and as a result increased video performance in high network latency conditions by an order of magnitude. We are currently working on adapting Google Earth to run under this system architecture, with the goal of the user being able to intuitively navigate through renderings of Ancient Rome in video glasses, without being hampered by any bulky equipment.

Peer instruction in computer science at small liberal arts colleges

Peer Instruction (PI) has been shown to be successful at improving pass-rates and improving retention of majors in large classes at large research-intensive institutions. At these institutions, students have been shown to learn from peer discussion in PI and both students and faculty have reported that they value PI in their classrooms. However, little is known about the effectiveness of PI in small classrooms at teaching-focused liberal arts colleges. This study evaluates results from seven lower-division classes and four upper-division classes taught at three different liberal arts institutions using PI. In these classes, PI experienced similar success as that reported at large-research intensive universities, both in terms of student learning from peer discussion and from student attitudinal surveys. Most notably, of 137 surveyed students, 91% recommend more faculty use PI in their classes.

Leveraging open source principles for flexible concept inventory development

Concept Inventory (CI) assessments, which target high-level learning goals, have proven highly valuable for higher education research. These assessments have helped to evaluate pedagogical practices among individual instructors, both within and across institutions, and have hence elevated the level of discourse on education within the community. The success of CIs in physics has inspired similar developments in computer science, with a few CIs now developed for computer science courses. However, the development of a CI typically follows a burdensome process, requiring a significant investment to produce a single CI that may be difficult to deploy due to institutional curricular differences. Furthermore, as our field continues to be shaped by technological advances, a path to faster, more modular CI development is critical. This paper proposes an alternative CI development model and continues the discussion within the community about the need for, and path to, concept inventories throughout the computer science curriculum. Specifically, we explore the implications of an open collaboration system for CI development that would mimic the principles common to open source software communities, which have regularly demonstrated their ability to produce high-quality results.

A Remote I/O Solution for the Cloud

With applications increasingly moving to the cloud, it is becoming common for an application to be separated by the network from the I/O devices with which the user is interacting. Currently this requires modifying the application to receive user input from the network rather than the device. We present a new I/O architecture in which the device driver is split into two parts, with the network between them. This architecture makes the network invisible to both device and application, allowing both of them to work unmodified. Our architecture also supports transformation modules, each of which comes in a pair that operates on each side of the network. Via these module pairs, the resulting system is capable of supporting the modification of the I/O stream in a variety of ways to compensate for the network, while remaining transparent to the application.

Browser Feature Usage on the Modern Web

Modern web browsers are incredibly complex, with millions of lines of code and over one thousand JavaScript functions and properties available to website authors. This work investigates how these browser features are used on the modern, open web. We find that JavaScript features differ wildly in popularity, with over 50% of provided features never used on the webs 10,000 most popular sites according to Alexa We also look at how popular ad and tracking blockers change the features used by sites, and identify a set of approximately 10% of features that are disproportionately blocked (prevented from executing by these extensions at least 90% of the time they are used). We additionally find that in the presence of these blockers, over 83% of available features are executed on less than 1% of the most popular 10,000 websites. We further measure other aspects of browser feature usage on the web, including how many features websites use, how the length of time a browser feature has been in the browser relates to its usage on the web, and how many security vulnerabilities have been associated with related browser features.

A highly-extensible architecture for networked I/O

We describe a new distributed I/O software architecture to support remote applications interacting with local I/O devices based on the concept of a networked device driver. Our goal is both network transparency and high extensibility/ease of customization in support of the vastly different types of applications and devices that can benefit from remote I/O, especially relevant in cloud computing contexts. A networked device driver logically connects a device at one network end-point and an application at another, and allows the I/O stream between them to be modified by a set of pipelined transformation modules. Each transformation module comes in a pair, operating on each side of the network, with one side typically applying some operation and the other side applying a corresponding one (e.g., one that reverses the original transformation). Because of the paired nature of transformation modules, the system is capable of supporting the modification of the I/O stream in a variety of ways to compensate for network issues while remaining transparent to the application, and also results in a high degree of extensibility. This is achieved with a mostly user-level implementation that incurs a relatively low degree of overhead.

Performance Aspects of Data Transfer in a New Networked I/O Architecture

We present performance results of a new distributed I/O software architecture to support remote applications interacting with local I/O devices. The architecture emphasizes network transparency and ease of customization/extensibility in support of the vastly different needs of various applications and devices that can benefit from remote I/O. Networked I/O is achieved via a networked device driver that is split into two parts, one on each side of the network. An I/O stream that is sourced at one end and sinked at the other may be modified by a set of pipelined transformation modules. Each module comes in a pair, one on each side of the network, with one side typically applying some operation and the other side applying a corresponding one, such as encoding and decoding the format of the data or pausing and resuming the sending of messages. Because of the paired nature of transformation modules, the system is capable of supporting the modification of the I/O stream in a variety of ways to compensate for network issues, one of the key problems of remote I/O, while remaining transparent to the application. We show that even with an implementation that operates almost entirely at user level (i.e., outside the operating system), good levels of performance that are adequate for even high intensity I/O, both in terms of efficiency and throughput, can be achieved.