Blog #6: TurKit

Paper Title: TurKit: Human Computation Algorithms on Mechanical Turk

Authors: Greg Little, Lydia B. Chilton, Max Goldman, Robert C, Miller

Authors Bios:

Greg Little is a student at MIT. His personal website is glittle.org.

Lydia Chilton is a Computer Science grad student at University of Washington. She got undergraduate degrees in Economics and EECS and a Master’s degree in EECS from MIT.

Max Goldman is a graduate student at MIT CSAIL Labs. His areas of study are User Interfaces and Software Development.

Robert C. Miller is an Associate Professor in the EECS dept. at MIT. He is also the leader of the User Interface Design Group in the CSAIL Labs.

Presentation Venue: UIST '10 Proceedings of the 23nd annual ACM symposium on User interface software and technology


Summary:Paper Hypothesis: TurKit is a toolkit for prototyping and exploring algorithmic human computation. Mechanical Turk (MTurk) provides an on-demand source of human computation. This provides a tremendous opportunity to explore algorithms which incorporate human computation as a function call. 
General Summary: Amazon uses such a mechanical turk in which users are paid small amounts of money for completing the human intelligence tasks (HITS) like writing product reviews. However, various systems challenges make this difficult in practice, and most uses of MTurk post large numbers of independent tasks. TurKit provides an API written in JavaScript for accomplishing MTurk Tasks. TurKit is a toolkit for prototyping and exploring algorithmic human computation, while maintaining a straight-forward imperative programming style. The authors present the crash-and-rerun programming model that makes TurKit possible, along with a variety of applications for human computation algorithms. The authors also present case studies of TurKit used for real experiments across different fields. The crash-and-rerun model favors usability over efficiency, but does so at an inherent cost in scalability. So, one of the major limitations of this system is expensive space complexity.

Discussion:
Companies like Amazon.com, ebay, other e-commerce companies and companies that rely on customer ratings have been using mechanical turks for quite some time now. They have to stick to this procedure since these tasks require human intelligence and can not be achieved using computer algorithms. These systems have a drawback when it comes to scalability, but with modern processors and cheap storage solutions, that doesn't quite seem to be an issue.

Blog #5: A Framework for Robust and Flexible Handling of Inputs with Uncertainty

Paper Title: A Framework for Robust and Flexible Handling of Inputs with Uncertainty


Authors: Julia Schwarz, Scott E. Hudson, Jennifer Mankoff, Andrew D. Wilson


Presentation Venue: UIST '10 Proceedings of the 23nd annual ACM symposium on User interface software and technology

Summary:

New input technologies like touch, recognition based input such as pen gestures and next generation interactions all provide for more natural user interfaces. However, these techniques all create inputs with some uncertainty. Conventional infrastructure lacks a method for easily handling uncertainty, and as a result input produced by these technologies is often converted to conventional events as quickly as possible, leading to a stunted interactive experience. The authors present a framework for handling input with uncertainty in a systematic, extensible, and easy to manipulate fashion.

Using the various touch inputs from the user, a probabilistic finite state machine can be developed and used to solve the ambiguity in certain inputs. These inputs include scenarios when the user touch the area half-way between the two options/buttons. In this case, probabilistic finite state machine can be used to resolve the conflicting inputs.

To illustrate this framework, they present several traditional interactors which have been extended to provide feedback about uncertain inputs and to allow for the possibility that in the end that input will be judged wrong or end up going to a different interactor. The six demonstrations conducted by the researchers included tiny buttons that were manipulable using touch input, a text box that could handle multiple interpretations of spoken input, a scrollbar that could respond to inexactly placed input, and buttons which were easier to click for people with motor impairments. The framework supports all of these interactions by carrying uncertainty forward all the way through selection of possible target interactors, interpretation by interactors, generation of uncertain candidate actions to take, and a mediation process that decides in a lazy fashion which actions should become final.

Discussion:

The paper talks about an excellent idea that's now being put to use. Today, more and more devices have touch screens. Soon, hardware buttons and controls will be replaced by touch screens on all the devices. When the primary method of interaction is touch, it becomes vital to resolve such annoying issues. This research has already been incorporated in many vitual keyboards. On iphones and iPads, when the user types words, he/she often makes mistakes since the letters are placed very close to each other. However, the algorithm computes the probabilities and displays the character that is most often times that character that the user intended to use.

Blog #4: Gestalt: Integrated Support for Implementation and Analysis in Machine Learning

Paper Title: Gestalt: Integrated Support for Implementation and Analysis in Machine Learning


Authors: Kayur Patel, Naomi Bancroft, Steven Drucker, James Fogarty, Andrew Ko & James Landay


Author Bios:
Kayur Patel is a student at the University of Washington pursuing a Ph.D. in Computer Science. He is a part of the DuB group and is being advised by James Fogarty and James Landay. He was previously supported by the US Government in the form of a Natioal Defense Science and Engineering Graduate (NDSEG) Fellowship.


Naomi Bancroft is an undergraduate student in the Computer Science & Engineering department at the University of Washington. She is currently doing research with Kayur Patel and James Fogarty on Human-Computer Interaction. She is a pursuing a double major in Computer Science and Linguistics.


Steven Drucker is a Principal Scientist at the LiveLabs Group at Microsoft. He is working on user interaction and information visualization for web based projects. He is also an affiliate professor at the University of Washington. Previously, he was a researcher at Microsoft Research for 11 years. First, as a lead researcher for the Next Media Research Group in Microsoft for 6 years where he examined how the addition of user interaction transforms conventional media.


James Fogarty is an Assistant Professor in the Department of Computer Science & Engineering at the University of Washington. He is broadly interested in Human-Computer Interaction, User Interface Software & Technology and Ubiquitous Computing. My research focuses on developing, deploying, and evaluating new approaches to the human obstacles surrounding the widespread adoption of ubiquitous and intelligent computing technologies.


Andrew Ko is an Assistant Professor at the Information School at the University of Washington. His research interests include social and cognitive factors in software engineering, end user software engineering, user interface software and technology, and programming language design.


James Landay is an Associate Professor in Computer Science & Engineering at the University of Washington, specializing in human-computer interaction. From 2003 through 2006 he was also the Laboratory Director of Intel Research Seattle, a university affiliated research lab exploring ubiquitous computing.

Presentation Venue: The presentation of this paper took place at UIST '10 proceedings of the 23rd annual ACM symposium on User Interface Software and Technology in New York.

Summary Section:
Paper Hypothesis: The hypothesis of the paper is to test the Gestalt tool. The authors test if Gestalt allows developers to implement a classification pipeline, analyze data as it moves through that pipeline, and can easily transition between implementation and analysis. They have focused on 5 major points in this paper:
1) Providing Structure While Maintaining Flexibility
2) Appropriate Data Structures
3) Visualizing and Aggregating Examples
4) Interactive, Connected Visualizations
5) The "Gestalt" of Gestalt
How the hypothesis was tested: Gestalt is compatible with most of the modern development environments like Eclipse, Microsoft Visual Studio etc). They tested the hypothesis by evaluating bug finding in Gestalt. They compared the bug-finding performance for participants using Gestalt with a baseline condition similar to MATLAB.
Result of the hypothesis: The participants unanimously preferred Gestalt and were able to find and fix more bugs using Gestalt than using the baseline.
General Summary: This paper presents Gestalt, a general-purpose tool for applying machine learning. Gestalt targets developers, providing full support for writing code to specify  the series of steps in a classification pipeline. In supporting a wide range of classification problems, Gestalt generalizes the lessons of prior domain-specific tools.


Discussion:
Significance of the paper: This paper provides a great motivation for users to create their own individual example visualizations. Also, Gestalt's connected visualizations make it easy to compare the data, attributes, and classification results of users.
Faults of the work: Gestalt has several limitations like the point where its tasks has pipelines that could be run in real-time. This can be proven expensive for many learning problems. Also Gestalt is not very effective in finding bugs in unfamiliar code.
Interesting future work: Gestalt is an effective and a relatively efficient tool that can help developers understand relationships between the various steps in a classification pipeline and can generalize advances from prior domain-specific tools to provide general-purpose support.

Blog #3: Pen + Touch = New Tools

Paper Title: Pen + Touch = New Tools


Authors: Ken Hinckley, Koji Yatani, Michel Pahud, Nicole Coddington, Jenny Rodenhouse, Andy Wilson, Hrvoje Benk and Bill Buxton


Author Bios:

Ken Hinckley is a Principal Researcher at Microsoft Research. The thrust of his research is to enhance input vocabulary and incorporate the use of computational devices and user interfaces.

Koji Yatani got a PhD at University of Toronto. During the course of his PhD, he worked with Prof. Khai N. Truong on the Dynamic Graphics Project. He has now joined the Human-Computer Interaction group at Microsoft Research Asia in Beijing.

Michel Pahud has a PhD in parallel computing from Swiss Federal Institute of Technology. He now works at Microsoft Resarch.

Nicole Coddington is a Senior Interaction Designer at HTC. In the past she has worked as a Senior Interaction Designer for Microsoft. Her specialties inclue NUI Interaction Design.

Jenny Rodenhouse is a Designer working for Microsoft in Seattle. Her current position is Experience Designer II at the Xbox Interactive Entertainment Division in Microsoft.

Andy Wilson is a senior researcher at Microsoft. His research includes working on applying sensing techniques to enable new styles of HCI.

Hrvoje Benko is a researcher at the Adaptive Systems and Interaction Group at Microsoft Research. His research interests are related to novel surface computing technologies and their impact on HCI.

Bill Buxton is a Principal Researcher at Microsoft Research. He is an advocate of innovation, design, consideration of human values, capacity, and culture in the conception, implementation, and use of new products and technologies. His teachings, research, talks and papers reflect these ideals.

Presentation Venue: This paper was presented  at UIST '10 Proceedings of the 23nd annual ACM symposium on User interface software and technology.


Summary Section:
Paper Hypothesis: The paper advocates a "division of labor" based on authors explorations, between pen and touch. They derive a system from these tools where the pen writes, touch manipulates and the combination of both produces new tools.
How the hypotheis was tested: The authors designed a study with a paper notebook to research on how people work with physical tools and pieces of paper. They asked the participants to illustrate their ideas for a short hypothetical film by pasting these pieces on a notebook. Then, they observed behaviors that informed them about the specific gestures as well as design attributes of their system.
Result of the hypothesis: The authors have listed the results of eight participants (all right handed) and listed various attributes about their design from B0 to B9 in the paper. These attributes basically provided them with the data of how different users use the pen. For e.g. some "tucked the pen between their fingers" in order to move pieces of papers. Some held the pieces of clippings and stopped writing and some did both simultaneously. These naturally occurring behaviors of users, served to them as the foundation for pen+touch gestures for their system.


Discussion
The paper provides a very intelligent and a smart way of designing a particular system. The authors use an excellent method of observing  and video taping the natural tactics of users in order to pursue tasks related to the project. Also, the authors clearly show how the combination of both, pen and touch, can bring about a change in the designs of products.


Interesting future work:

Blog #2: Hands-on Math

Paper Title: Hands-on Math: A page-based multi-touch and pen desktop for technical work and problem solving

Authors: Robert Zeleznik, Andrew Bragdon, Ferdi Adeputra, Hsu-Sheng Ko

Author Bios:
Robert "Coach-Z" Zeleznik is a Director at the Research Department for the Computer Graphics Group (Brown University).

Andrew Bragdon is a Ph.D. student at Brown University, department of Computer Science. He is currently developing methods of making computing tasks more fluid, natural and cognitively lightweight.

Ferdi Adeputra is a former student of Brown University, department of Computer Science, currently employed at Goldman Sachs. His works apart from Hands-on Math include Code Bubbles, where he proposed a novel user interface that is based on collections of lightweight editable fragments (called bubbles).

Hsu-Sheng Ko is a former student of Brown University, department of Computer Science, whose publishes include FLUX - A Tilting Multi-Touch and Pen Based Surface, where FLUX is an interactive touch-sensitive tilting surface that can be used either as a digital presentation whiteboard and other interactive surfaces.

Presentation Venue: Proceedings of the 23rd Annual Symposium on User Interface Software and Technology (UIST 2010)

Summary Section:
Paper Hypothesis: The papers' core hypothesis is that if CAS (Computer Algebra Systems) tools were driven by direct, multi-touch manipulation and digital ink within a free-form note-taking environment, students and even scientists might learn and work more efficiently.
How the hypothesis was tested: The hypothesis was tested using a qualitative pilot evaluation and usability discussion of the prototype system as a whole. The hybrid tabletop solution that were designed by the authors in order to blend the whiteboard-like interactions with resizeable pages was tested using several gestures such as the bezel and folding gestures. These gestures helped them test the creation and deletion of pages, and to test page management within a page respectively.
Result of the hypothesis: The 9 participants that were recruited from the undergraduate student population of Brown University were very positive about the potential of the system. They all found the virtual papers to be as comfortable as natural paper, and they were also enthusiastic about the panning bar. The system was remarked as "cool" and was reported to be helpful to them to organize their pages into distinct regions.
General summary: The participants approved the potential of the system but they faced a few difficulties in operating certain features of the same.

Discussion:
Significance of the paper: This paper has huge significance on society as a whole as it provides features that can lead to a minor technological advancements in the usage of technology in order to solve and tutor math. One of the most important features it provides is it uses virtual paper, which if used even to a mediocre extent can save a lot of paper and thereby a lot of trees in the environment. The system also contributes novel bi-manual and gestural techniques for managing and writing on virtual note pages in addition to direct manipulation techniques for algebraically transforming mathematical expressions.
Faults of the work: The participants faced various problems which included difficulties in: page creation and deletion, pinching, multitouch dragging and zooming.
Interesting future work: This technology can be implemented in touch devices such as touch phones and touch pads in order to make mathematical computations easier and faster.

Paper Blog #1: Imaginary Interfaces

Paper Title: Imaginary Interfaces: Spatial Interaction with Empty Hands and without Visual Feedback

Paper Authors: Sean Gustafson, Daniel Bierwirth and Patrick Baudisch

Author Bios: Sean Gustafson is a PhD student at the Human-Computer Interaction Lab of the Hasso Plattner Institute in Potsdam, Germany and is supervised by Dr. Patrick Baudisch. His main research interests are in understanding the fundamentals of new interaction paradigms, particularly gestural and eyes-free use of portable technology.

Daniel Bierwirth worked on a variety of projects for software visualization and scientific visualization. Some of his past job positions were at startup companies such as Software Diagnostics GmbH and Dynardo GmbH.
Patrick Baudisch is a computer science professor at Hasso Plattner Institute Potsdam, Germany. He is also the chair of the Human Computer Interaction Lab at the institute. His research focuses on the miniaturization of mobile devices and touch input.

 Summary: The paper "Imaginary Interfaces" discusses the idea of having screen-less devices that allow users to perform spatial interaction with empty hands and without visual feedback. The paper discusses spatial interaction in an invisible 2D space compared to conventional computer systems where objects have a specific location on a display device. There are several examples in the paper that shows visual "feedback" taking place in the user's imagination. For e.g. there is an example where Karl (user) uses an imaginary interface to graphically represent his stock options by forming an 'L' with his left hand. The paper also discusses many ideas on ways to successfully build imaginary interfaces on wearable computing, gestural input, mobile computing and spatial interaction.

Discussion:
Significance of the paper: The idea of spatial interaction, if implemented correctly, can lead to a big technological advancement. The concept of creating an imaginary interface can be really advantageous in many ways. One of the biggest advantages of this technology would be the concept of screen size and pixels would vanish. Although a lot of things would be dependent on the user's short term memory, the idea can prove to be very effective for small tasks such as analyzing small data. People can use this technology in their everyday lives and increase their work efficiency. The technology can bring advancements in mobile technology and can also decrease E-Waste substantially!

Blog #0: On Computers

Reference: Aristotle, On Plants, The Revised Oxford Translation, Edited by Jonathan Barnes, Vol. II

Author Bios: Aristotle was born in Stagira in north Greece, the son of Nichomachus, the court physician to the Macedonian royal family. He is said to have written 150 philosophical treatises. The 30 that survive touch on an enormous range of philosophical problems, from biology and physics to morals to aesthetics to politics.

Summary: In "On Plants", Aristotle** discusses and justifies how plants should be considered to have souls like animals and humans. He starts by stating how life in animals is clearly manifest whereas the same in plants is not clearly evident. He tries to build a relationship between science and emotions by trying to relate the process of photosynthesis and fruit bearing within plants to them asserting a sensation of pleasure and desire. He provides many similar comparisons between the physical structures of plants and animals and how they can be called similar.

Discussion: While reading "On Plants" I learned how Aristotle used abundant examples to back his statements about plants possessing a soul. Relating this idea to computers and pondering upon whether they have a soul or not is a different and unique thought process. Although the arguments and points described in the book are extremely detailed and descriptive, I think it is not a correct way to prove a point. However, even if we think about computers in that perspective, I believe computers do not have soul. Firstly there is a big difference between living things and machines. There are many properties that plants sustain that can bring them into consideration for probably possessing a soul (for e.g. the sunflower exhibits heliotropism) but computers are machines that contain knowledge that is inputted by the developers/users. If information from all the books and libraries were to be inputted and stored into a computer, and then if that computer was made to answer questions asked by humans; it doesn't mean that the computer has a brain and that it is answering questions by itself (assuming that its answer does make sense)! It simply means that there is a simple search mechanism being implemented and the results being displayed by a machine. This can no way be related to a computer possessing a soul.