From Geek to Sleek: Integrating Task Learning Tools to Support End Users in Real-World Applications

Authors

Aaron Spaulding, Jim Blythe, Will Haines, Melinda Gervasio


Artificial Intelligence Center
SRI International
333 Ravenswood Ave.
Menlo Park, CA 94025
{spaulding, haines, gervasio}@ai.sri.com


USC Information Sciences Institute
4676 Admiralty Way
Marina del Rey, CA 90292
blythe@isi.edu

Summary
These researches combined forces to create ITL, an integrated task learning system.  Basically it aims to provide a more user friendly script creation for the average end user.  They wanted to make something that was not nit-picky as macro recorders as well.

They integrated their technology with two currently existing military software applications called LAPDOG and Tailor.  Initial versions were disliked by users and the product was updated to adjust to their suggestions.In addition to letting users specify a particular set of scripted steps, the program let the users edit, copy, and paste scripts as well. In the end, they found that their product does in fact increase productivity because it automates repetitive tasks for the users.

Discussion
My favorite line was "this format was immediately derided as 'a bunch of geek [stuff]' by our users". I may have actually chuckled out loud a little to see those brackets in a scholarly paper.  Also, considering the simplicity of the study's topic I understand why this was labeled under the "short" papers list.

Pulling Strings from a Tangle: Visualizing a Personal Music Listening History

Authors


Dominikus Baur
Media Informatics, University of Munich
Munich, Germany
dominikus.baur@ifi.lmu.de


Andreas Butz

Media Informatics, University of Munich
Munich, Germany
andreas.butz@ifi.lmu.de


Summary

Several methods for helping music listeners create a playlist are currently implemented.  Most mp3 players utilize ID3 tags which contain album, artist, year, genre, etc information.  While this information allows the user to quickly utilize and find particular songs, playlist creation requires intimate knowledge of the actual music that lies the behind the ID3 tags.  Others such as Pandora, have utilized machine learning techniques to categorize songs and base future suggested songs off of the machine's recommendations.  Still others, such as Last.fm, study past listening histories across thousands and thousands of users and based on the similarities between those and the user's listening history suggests future songs.

To obtain the data needed for the project, they used Audioscrobbler, Last.fm's tracking software. Using this data, they created three different types of visualizations.  First is the Tangle, which is a global view of the listening history.  Outer loops represent anomalies or songs that were listened to without correlations to other songs.  Songs that are close to other songs are relatively similar and thickness of a line represents how often a particular sequence of song was listened to.

The second visualization algorithm is strings,  which shows, in a very ordered and organized way, past listening sequences.

The third visualization is called knots and it shows the user which songs are present in several strings to give the user a better visualization of which songs they most listen to.

Users can then determine paths that their music should take along with specifying a "novelty" factor which determines how unusual the path between waypoints is.


Discussion
This sounds like it would be a really cool but novel idea for the music culture. It would be equivalent to the visualizations that iTunes and Windows Media Player offer.  They're really cool to look at but no one really uses it more than a couple times.  I could be wrong because I haven't exactly used it and unlike those visualizations offered by Windows and Apple, Tangle actually offers some practical usage.  The 1,000 song limit is a bit of a problem though.

Parakeet: A Continuous Speech Recognition System for Mobile Touch-Screen Devices

Authors

Keith Vertanen and Per Ola Kristensson
Cavendish Laboratory, University of Cambridge
JJ Thomson Avenue, Cambridge, UK
{kv277, pok21}@cam.ac.uk


Summary
This paper explores the possibility of speech recognition as a way to input text on a touchscreen device.  Speech recognition has generally been blocked from entering the mobile phone arena because of the processor resources required to effectively translate from speech to text.  Not only does it normally require far more resources than the processor can spare, the result is usually very erroneous (ie Google Voice voicemail transcriptions).  However, despite the setbacks speech to text is very attractive as a text input method because users are already familiar with speaking so no training is involved and users can typically speak up to 200 wpm which far surpasses even the best typists (especially mobile typists).

The researchers built Parakeet to address this issue. When designing it they followed 4 design principles.  First, they were to avoid cascading errors involving phase offset.  They did this by using a multi-modal speech interface.  Second, they needed to exploit the speech recognition hypothesis space.  They did this by offering the user several choices from several recognition hypotheses.  Third, they must implement efficient and practical interaction by touch.  They did this by implementing an interface that only required touch so that it could be used with or without a stylus.  Lastly, the design should support fragmented interaction.  In other words, the device should anticipate the ADD tendencies of its user.  They did this by offering text predictions and alerting the user audibly that their speech is done translating to text.

Users used the device both while seated indoors and walking outdoors to test the effectiveness and mobility of their product.  They were given a fixed set of sentences between 8 and 16 words in length to say in 30 minute trials.  Once they said a sentence they were to try and correct the sentence using Parakeet's correction interface before moving on to the next sentence.  On average, participants completed 41 sentences inside and 27 sentences outside during the 30 minute trials.

Results indicated that participants managed an average of 18 wpm indoors and 13 wpm outdoors while utilizing correction.

Discussion
I don't really know what the average typing speed of phone texters is but I would imagine that 18wpm is a little on the slow side for texters, especially those that text frequently.  Obviously, this rate would be vastly improved if the translation algorithm was more accurate, but even the almighty Google is far from achieving that feat.

“My Dating Site Thinks I’m a Loser”: Effects of Personal Photos and Presentation Intervals on Perceptions of Recommender Systems

Authors

Shailendra Rao
Stanford University
shailo@stanford.edu


Tom Hurlbutt
Stanford University/ Intuit
hurlbutt@cs.stanford.edu


Clifford Nass
Stanford University
nass@stanford.edu


Nundu JanakiRam
Stanford University/ Google
nundu@google.com


Summary


This study explores the effects that bad recommendations from recommender systems can have on the individuals that use those systems.  For example, if someone's TiVo mistakenly thinks that they are gay and recommends television programs that would appeal to those that are gay, then the user might erroneously watch more "guy" programming to compensate for their miscategorization.

The situation that is explored in this paper is online dating.  People will lie about themselves, omit information about themselves, or submit misleading pictures of themselves to present their "ideal" self.  Giving false information like this only skews the system and matches users with people who differ substantially from what they were expecting.  This leads to frustration for both parties involved and now the dating site cannot effectively carry out its purpose or end goal.

The study used a new algorithm called MetaMatch, which took the answers to a set of personal questions and used them to purposefully suggest undesirable matches.  After receiving the matches, the participants rated their recommendations and answered a post questionnaire.

During the questionnaire, half of the participants had their submitted picture posted next to the questions as they answered them and the other half had no picture.  Also, half of the participants were shown 4 possible (undesirable) matches  after each set of 10 questions and the other half were only shown matches after they finished the entire questionnaire.

The results of the post questionnaire are represented in the graphs above.  The study found that displaying a personal photo can have a stabilizing effect amongst participants. Having an element of self-reflection during the questionnaire tended to make the participants not change their answers to the questions to compensate for their bad matches. Also, intermediate recommendations tended to make the user change how much they approved of the bad match at the end of the questions.

Discussion
I thought this study was one of those studies that proves something that you already suspect.  It would be interesting to see what online dating companies do with this information or with more indepth studies.  I do feel like only testing 56 people was a bit small for the user study though. I would like to have seen a more exhaustive study that included thousands of participants.

The TeeBoard: an Education-Friendly Construction Platform for E-Textiles and Wearable Computing

Authors
Grace Ngai, Stephen C.F. Chan,  Joey C.Y. Cheung, Winnie W.Y. Lau
{ csgngai, csschan, cscycheung, cswylau }@comp.polyu.edu.hk

Department of Computing
Hong Kong Polytechnic University
Kowloon, Hong Kong


Summary

Wearable computing and e-textiles is an emerging field in contemporary fashion. However, existing technology is so complex that it is neither affordable or feasible for the average hobbyist or consumer.  The solution for this problem, presented by the above researchers, is the TeeBoard, a constructive platform that aims to "lower the floor" for e-textiles and explores the possibility of integrating the technology into the education field as a teachable tool.

For e-textile technology to be more widely adopted as an educational tool it should be robust for everyday use and rough handling, it should have large error tolerance, and a shallow learning curve. The current state of e-textile technology does not yet satisfy these requirements.

The researchers set out several goals for their project in addition to solving the above listed problems: 1) capabilities of the materials used should be as close to their electrical equivalents as possible, 2) the construction platform should be able to be used by users of all skill levels, and 3) the platform should be easily reconfigurable and debuggable.

The TeeBoard was the solution they came up with, a wearable breadboard that could easily configured to the user's wishes.  The shirt uses a Arduino Lilypad microcontroller for processing of signals and a conductive thread instead of wires to achieve the robustness necessary for everyday wear and tear.

The shirts were used with several test groups.  Groups were given simple tasks and then incrementally harder tasks to test the effectiveness of the product.  The results showed that the TeeBoard did in fact meet its requirements and students were more interested in computational and electronic designs.




Discussion
This is an interesting concept for more interactive applications of electronic theory.  I think the main problem with the A&M program is that application is not stressed in a tactile way. Granted we do use breadboards and circuits to illustrate concepts, but rarely are the concepts illustrated in a way that it would be used in the real world.  I would love to see more interactive solutions such as the one in this paper in the educational system.

The only problem I saw with the user study was that it was tested amongst college classes that already had significant background in electronic theory. I would have liked to see them test this product amongst high school settings where students are not as familiar with electronic theory.

PrintMarmoset: Redesigning the Print Button for Sustainability

Authors

Jun Xiao
Hewlett-Packard Laboratories
Palo Alto, CA USA
jun.xiao2@hp.com


Jian Fan
Hewlett-Packard Laboratories
Palo Alto, CA USA
jian.fan@hp.com

Summary
This paper focuses on Sustainable Interaction Design, design which aims to alleviate the problems associated with dwindling natural resources by shifting from production of new products to production of sustainable growth products. Specifically, they focus on redesigning the print button to minimize erroneous prints and other various paper wastes and in turn, promote sustainability.

Users in the study were studied to reveal what and how often they printed and the results were critically examined to draw conclusions. The found that users printed information that was digitally available because it saved time(in the case of printing a map vs using a GPS), it was easier to user(in the case of printing a pdf instead of viewing it on screen), it signified importance(in the case of printing a website instead of emailing a link), and the information was volatile and needed to be preserved(in the case of news articles being constantly updated).  These cases were generalized to say that printing served four main functions: display, delivery, sharing, and storage.

While this paper did not aim to resolve these problems, it did attempt to minimize the problem of erroneous or unnecessary prints. People reported printing out webcontent because it was necessary but only to find their printout riddled with web advertisements or malformatting which rendered the printout useless and therefore thrown away as waste.

Their goal was to create a solution that required 1) require neither developers to modify existing web sites nor users to change existing print flow; 2) require the least amount of user input effort, if not zero; 3) offer flexibility that allows users to choose what to print in addition to pre-defined filter or template; 4) maintain a history of print activities for collecting and organizing web clips; 5) raise awareness of the tool among people and build a “green” community around it.

The resulting solution was PrintMarmoset, a Firefox extension that allowed the user to better select what needed to be printed.  The user would click the icon for the extension, "stroke" through webcontent that was unnecessary, and then click the icon again to print what was selected.

People commented that PrintMarmoset is WYSIWYG and the UI was easy to use. The authors commented that the tool did not satisfy all users but that that was an impossible task to accomplish.

Discussion
This seems like an awesome idea to be honest.  Many times have I wanted to print a website only to see that the advertisements have completely ruined the layout of the print or have offset the layout to produced unneeded blank pages with just the website printed at the bottom.  These pages are then thrown away as a wasted print and paper is subsequently unnecessarily wasted.  I'll have to try this extension out to see just how user friendly it is, but the concept itself seems hopeful.

Emotional Design

"The problem is that we still let logic make decisions for us"

Sentences like this in the prologue of the book kind of scared me going into it.  However, I did enjoy this book more than his previous work. It gave more insight into why certain technologies succeed over others despite offering the same or even less capabilities than their competition. I also enjoyed how spelled out the processes of what was happening when a person is introduced to a new product.

Honestly, I left reading the book with one desire in mind: I'd love to visit this guy's house. He kept referencing all the little things he has in his house and i think it would be so cool to just get to mess with it all one day.