User-Oriented Document Summarization through Vision-Based Eye-Tracking

Summary:
In this paper, the authors present their new algorithm for document summarization. Their prototype uses eye-tracking to determine which words in a document are focused on the longest by a reader. The algorithm then predicts how long users would focus on other words based on semantic similarity. Basically, their algorithm predicts what sentences the user is most likely to focus on, and then ranks them in decreasing order so as to present the "most important" information for that specific user first. What sets this apart from other summarization techniques is the fact that the documents are broken down into components (word by word) that are then combined to rank sentences and paragraphs.

Eye tracking sample. This person looks at weird things.

While testing their new algorithm, the authors continually beat MS Word AutoSummarize and MEAD (an open source summarization package) in both Precision and Recall. In the future, they hope to improve their algorithm so that documents a user has never read can be successfully summarized (currently, it only works if the user is going back over something previously read).

Discussion:
I didn't know that automatic summary technology even existed. It seems to me that the authors are moving in the right direction by having summarization be user-oriented instead of generic. Everyone reviews differently and focuses on different things; a software that can learn your style would be very valuable. Their future work hinted at creating a summarization tool for things that the user hasn't even read before. I could use that on my next blog!

Do You Know? Recommending People to Invite into Your Social Network

Summary:
The explicit definition of who is your friend within a social network is created by you. If Facebook forced you to be friends with everyone in your town, or school, or family then odds are you wouldn't want to use it. However, Facebook and other sites do offer suggestions as to who you may want to be friends with. Facebook will sometimes even give you a brief reason, such as the fact that you both went to the same high school.

In this paper, the authors discuss their recommendation widget, "Do You Know?". It combines information from the social network and outside of it to make recommendations that are viewed one at a time with the reasoning behind the choice. DYK is available for IBM's employee directory (all of the authors work for IBM), and 6287 users were logged during its initial test study. Overall, employees were happy with the tool, and provided feedback to the designers. The main point of question was the "No Thanks" button, which served as a way to permanently remove a suggestion. Some users were unsure as to just what it did (i.e., did it tell the person you rejected them?), and others never even noticed it. You can see a sample of DYK below.

Discussion:

Why is IBM concerned with friends, anyway? It seems strange to me that they would put an emphasis on the connection between employees while in the workplace. Now people can go through DYK while playing Mafia Wars 2.

But seriously, the idea behind DYK is cool. If people who have worked on projects together or have similar interests within the company can find each other and communicate, then they are more likely to feel a tie to IBM and fellow employees. Also, they may be able to collaborate on future projects together by finding compatible team members. Other social networking sites could learn something from this idea. Facebook likes to pull recommendations just from FB info, hence I get suggestions for people who go to A&M. Thanks Facebook! I definitely know that one girl from the 40,000+ who attend school here.

PrintMarmoset: Redesigning the Print Button for Sustainability

Summary:

Jun Xiao and Jian Fan from HP Laboratories conducted a study aimed at reducing the amount of wasted paper that results from web printing. Their browser add-on, named PrintMarmoset, allows users to remove web content before printing. Their field of interest is Sustainable Interaction Design (SID). SID focuses on creating sustainable and evolving products and technologies. Where most design is interested in how one person uses a system, SID is interested in how everyone can use a system with the smallest impact on the environment or resources. With that in mind, the authors studied the printing behaviour of computer users. They mention that 50 TRILLION pages were printed in 2007 to show that SID has could and should be applied. By conducting interviews, the authors found that people print things to signify their importance, and because it's faster than transferring data electronically (their example is a map versus programming GPS).





To combat the excessive waste associated with printing things online, the authors developed PrintMarmoset. They tried to make their add-on as simple to use as possible, thereby minimizing the chance that people wouldn't use it because it frustrates them. This also promotes SID, by incorporating it in an unobtrusive way. Users simply select parts of the website that they do not want to print, and then whatever is left will be printed. PrintMarmoset is a WYSIWYG tool, and as such users in the study preferred it to current printing methods.

You should definitely click the image.

Discussion:
Have you ever printed an email and a second page comes out with just an address on it? Or pages are covered with ads that, even though you click 'Print Selection', you end up having pages and pages of? PrintMarmoset seems like a tool that was needed years ago. So why didn't anyone think of it? I think the authors are right when they say that people think paper is plentiful. It seems like everyone has this mindset that you can print things a thousand times at maximum size and spacing, and therefore wouldn't need a tool to reduce waste. Just think about Reed... that recycle bin is filled with crap that people didn't want to print. I hope that the authors will push for their add-on to be available in all browsers.

The Application of Forgiveness in Social System Design

Summary:
Asimina Vasalou, Adam Joinson (both University of Bath), and Jens Riegelsberger (Google, Ltd., U.K.) are interested in applying the concept of "forgiveness" to social systems. Online communities offer many ways to control bad users, such as bans, ratings, and filters. The problem arises when good users have a lapse in judgement or make an honest mistake, and are reprimanded for it just like intentional abusers of the community. To prevent lasting harm, the authors suggest a system for social networks and communities that promotes forgiveness for people who would normally have a high record or popularity.

The authors define forgiveness as follows:

"Forgiveness is the victim’s prosocial change towards the offender as s/he replaces these initial negative motivations with positive motivations."

They then list and briefly detail seven factors that can help victims move beyond an offense (I won't go into detail, but you can read about it!). These factors are Offense Severity, Intent, Apology, Reparative Actions, Non-verbal Expressions, Dyadic History, and History in the Community. As is true in real life (aka life outside of computers), forgiveness is not a guarantee. To mirror this, the authors suggest that any social system choosing to implement forgiveness should consider three key things (they apparently love lists): forgiveness isn't mandatory or unconditional, and it doesn't repair trust or remove accountability. They stress that incorporating forgiveness allows online communities to instill a sense of empathy in members, and that both offenders and victims are given the chance to recover their community status. If communities incorporate a system of forgiveness as well as they have systems of punishment (referred to as 'reparative design'), then authors think everyone wins.

Discussion:
After reading this paper, I'm surprised that communities don't already incorporate reparative design. While playing Counter-Strike:Source with friends on the PC, I often see players banned after first offenses. When I was an Admin for a server, I noticed that there weren't any warnings except temporary bans. Sometimes people really get into things, and will accidentally break a rule (such as no cursing). Do servers forgive them? Nope. They own them. Then the offender is mad, the victims could care less, and the lesson learned is that the server or community sucks. I think that future work could study the application of reparative design in certain forums or servers. It's one thing to propose an idea; it's another to see if people will use it.

Learning from IKEA Hacking: “Iʼm Not One to Decoupage a Tabletop and Call It a Day.”

An IKEA chair before and after hacking.

Summary:

Daniela Rosner and Jonathan Bean (UC Berkeley) are interested in IKEA Hackers- people who use IKEA products to make new, customizable things. Because these hackers share their ideas and creativity online with others, they are promoting CHI via the DIY community.

IKEA is the source and inspiration for these hackers for a few reasons. First, IKEA furniture is aesthetically pleasing to them, thus allowing them to create art along with functional things. Second, IKEA is cheap. Some of the hackers surveyed treated IKEA like a hardware store for raw materials instead of full products. This also lets hackers feel more at ease about cutting pieces up and going against instructions, thus allowing them to make just what they need for a given place or problem.

In terms of using technology, IKEA hackers often place their concepts and changes online. Many websites, such as Instructables.com and IKEAHacker.com, offer opportunities for hackers to share their new instructions and techniques with others. The sense of community online lets the hackers feel like they belong and keeps them motivated to create. The authors note that people are merging computer terms (such as hacking and programming) and physical items (IKEA furniture) to create a wholly unique CHI experience and culture. In their words,

"...DIY culture is moving the workshop from the garage to the
web forum."

Discussion:
I'm a member of Instructables.com, and can testify to the craziness and awesomeness of tweaking products and designs to make new things. I have a bookshelf made out of Tetrominoes in my bedroom that I found a guide for of the site. With the exception of the creepy gyno chair that sits on the first page of this article, I found it to be really cool. Now as for the people interviewed... they seem a little out there. But what artist or sculptor isn't?

An Exploration of Social Requirements for Exercise Group Formation

Summary:
In this article, Mike Wu, Abhishek Ranjan, and Khai N. Truong (all from the University of Toronto) explain their findings on how people find exercise partners. They used an online survey of 96 people, followed by two focus groups of 12 people. The authors wanted to find out which characteristics of working out with others could be translated into the design of applications and websites devoted to bringing people together for exercise. Their results can be broken down into the following:

• Most people have or look for exercise partners
• People generally know their partner prior to working out
• Most people will not exercise alone
• People are willing to share different information to find partners (figure 1)
  
  

From these findings, the authors suggest that social networks geared towards helping people find exercise partners should have a few characteristics. First, they must allow people to collaborate about what they like and when they are available. Second, users should be allowed to update their available times constantly, instead of setting rigid schedules on calendars.Third, users should be able to vary what personal information they share with other people in the network. And finally, users should be able to choose their partners based on familiarity instead of being matched with strangers.

Figure 1. How willing people are to share information

Discussion:
I found this study to be interesting and true. I know that I have a hard time making myself go to the Rec unless someone is going with me. Even then, it takes a while to hash out schedules and communicate times. Generally, it's texting that gets everyone together, so having a mobile application based on these findings would be useful and practical. I would expect that to soon be developed as future work (the author's expect it, too). I wonder if the findings would be different for Americans over Canadians...

Social Computing Privacy Concerns: Antecedents & Effects

Summary:

Oded Nov (New York University Polytechnic Institute) and Sunil Wattal (Temple University) wrote this paper to discuss their findings concerning the social aspects of privacy in social computing communities. These communities include blogs, forums, wikis, and social networking sites such as Facebook. They wanted to answer two main questions. First, do social norms and trust in fellow members have an impact on privacy concerns? And second, does the social network structure and length of membership play a part?

For their study, the authors looked at photo sharing on Flickr. People who use Flickr can choose to share their photos with a variety of groups ranging from everyone to nobody. 192 people agreed to share their information via Flickr's API, and from this the authors based their findings.

results = lol

In the end, the authors found that users will choose more restrictive privacy settings and limit what information they share because they do not trust other community members and because they follow that community's norms.

Discussion:

The "lol results" above could easily have been replaced with "who cares results" as far as I'm concerned. This research was mostly just a conglomeration of references to the things that other people had done combined with some overly complex statistics. I could really care less about how people on Flickr react to privacy settings and not trusting others. Seriously... I thought this paper was going to rock. Instead, it took me four hours to read it and write this blog about it because I immediately realized that I would rather watch Weather Scan over and over than get to page three. Maybe it's just me... or maybe they had an awesome idea and then lost it somewhere. Or maybe the "lol results" are what they wanted to show.

“It Feels Better Than Filing”: Everyday Work Experiences in an Activity-Based Computing System

Summary:
In this paper, Stephen Voida (University of Calgary) and Elizabeth D. Mynatt (Georgia Institute of Technology) studied the use of an activity-based computing system named Giornata. Activity-based computing boils down to designing interfaces that allow users to organize their tasks and information in a personal way. This differs from typical desktop computing because users can tag/manipulate items and collaborate instead of simply placing items in hierarchical files and share the files without good descriptions. In order to better understand the potential benefits or hindrances of Giornata, the authors chose to create a full-featured system that was used over an average span of 54 days by the 5 focus users (two faculty members, two grad students, and an industry member).

Giornata was made for Mac OS X. Click the screenshot for a suck-free view.

Once opened, Giornata provides a virtual desktop for each activity, wherein all work pertaining to that unique activity is carried out. This can be compared to having multiple desktop tabs, where each desktop is devoted to something different. Tags can be applied to each activity, and documents that span different activities inherit all relevant tags. These tags can be searched over for easy recovery of information instead of just file names. Users were also able to collaborate with each other by sending activities via email directly from their activity's desktop.

The users reported generally positive experiences using Giornata. Because they were allowed to freely use the program without restrictions on time or features (as in a typical user study), each user developed their own habits of information manipulation and activity creation. Tagging was used by every person on at least one of their activities, but the users found it to be of limited use in the short-term, although they stated that they could anticipate long-term benefits. In addition, users kept all of their needed information on their virtual desktop instead of archiving it in a file system.

The authors concluded from their study that customizable visualized data is a benefit to users, allowing them to work freely instead of through a hierarchical structure. Also, they believe that having activity-based storage and areas allow people to more naturally organize their work as they would without using a computer.

Discussion:
I had never heard of activity-based computing prior to this paper, and I now find the topic to be pretty interesting. Being able to visually compartmentalize my work would be great. I have a habit of placing everything I need in a file structure, and then either forgetting how things are related or forgetting what items are due when. Giornata seems like it would help me with both issues. I would be interested in seeing this research expanded to include tools that are specific to certain activities, such as writing research papers and maintaining all needed information. Such an interface would be ideal for applications such as PowerPoint and Publisher as well.

The Inmates are Running the Asylum (Part 1)

Summary:
Alan Cooper wrote The Inmates are Running the Asylum for the sole purpose of exposing software programmers and their corporations to the lack of attention they pay to design. Cooper frequently uses the term "interaction design" to mean a design that is easy for users to understand and use to the point that they will choose a product because of it. This is a fundamental choice that programmers must make before designing products instead of simply tacking on an interface to try and mask their confusing code. The problem is that programmers can understand what they make (after all, they made it) and therefore cannot see how customers would have an issue understanding it as well. Cooper points out that people in the software industry think more like computers than humans, and therefore have detached themselves from users who aren't experts in the field of computers.

Cooper breaks these users into two categories: Survivors and Apologists. Survivors are those people who cannot stand software because they cannot understand software. They frequently feel stupid and embarrassed by their inability to get products to work the way that they need them to, and will silently stick it out because they don't think there is a better way available. On the other hand, Apologists think that software is the bee's knees. Cooper (rightfully so) compares them to hardcore political party advocates, who will declare their party the ultimate good and ignore any and every evil. They are loyal not to good products, but to products in general. Therefore, they work hard to master difficult programs because they feel it is their right and privilege to do so. Like Survivors, they cannot imagine a better way to interact with software because they have developed a digital case of Stockholm Syndrome.

Cooper gives many examples showing that failing to put user needs and experience as the top priority has killed products and companies. At the end of Chapter 6, he states that engineering methods cannot solve the problem because they are the problem. Homo logicus creates a coding masterpiece, yet Homo sapiens cannot use it. Without designing for humans, humans will never enjoy their masterpiece.

Discussion:
Seeing as how I'm already done with the first half of the book, it's obvious that I found it really interesting. Cooper's writing style meshes well with his ideas of interaction design, and before you know it you're thirty pages in. What surprises me most about this book so far is the fact that most of the companies and products he mentions are things that I have never even heard of. This is a true testament to the power of his statements. These dead companies piled on features instead of refining design, and suffered for it. In my opinion, the lack of user-centered design helped lead to the burst of the tech bubble, with so many promising companies collapsing in on themselves. Apologists built them up, but in the end the companies had nothing to offer the real consumers who needed new technology without knowing it. So far, a great book.

How Well do Visual Verbs Work in Daily Communication for Young and Old Adults?

Summary:
In this paper, Xiaojuan Ma and Perry R. Cook (both from Princeton) analyze the different ways that verbs can be visually displayed, and how these displays are conveyed to both young (20-39) and old (55+) adults. The four visualizations in question are a single image, a panel of four images, an animation, and a video clip. They chose 48 frequently used verbs from the British National Corpus to visualize for their research. Note that verbs are harder to convey via images as compared to nouns, as most nouns represent a single, tangible thing.

The four visualizations for "work".

The images for the verbs were taken from tagged web pages. The author recruited raters to give them feedback as to watch images best depicted each verb, and then chose the top four. The animations were taken from a sight specializing in verb animations, and the videos were recorded by the authors themselves.

After comparing the four representations to a group from each age population, the authors found that young adults performed better for all visual modes. The values of responses for both these groups were considered by WordNet score. This scoring is done on a six-point scale based on distance from the target verb. An exact match between the verb and the user's guess is awarded 6 points, a synonym is awarded 5 points, and an irrelevant guess yields 1 point.

WordNet score results for the two groups

The authors also found that the verbs best recognized all shared certain visual characteristics in common. Some of these include simple backgrounds, limited visual effects, and a limited use of symbols (such as a heart or a question mark to represent the verb). They also make note of the fact that different gestures have different meanings to people of different cultures and age groups, and therefore the visualizations should be examined for universality. In the future, the authors hope to apply their visual designs to help people with Aphasia (a disorder categorized by understanding written and spoken language).


Discussion:
I thought this article was interesting. I didn't actually see the purpose for the research, however, until I read where they were interested in helping people with Aphasia. With that in mind, it casts the research in a whole new light. By visualizing verbs, people who have difficulty speaking and writing can still communicate with others, which is of obvious importance. I think that future research should be done to with actual people suffering from Aphasia so that the benefits to them can be directly determined. Maybe an entire, universal language could be created based on visualizations, something of great benefit to anyone attempting to communicate across language barriers.

Correlating Low-Level Image Statistics with Users’ Rapid Aesthetic and Affective Judgments of Web Pages

Summary:
In this article, the authors were concerned with how well websites evaluated based on decomposed low-level images compared to what actual users thought about the pages themselves. Users were asked to base their decisions on four main design dimensions:
Attractiveness, Pragmatic Quality, Hedonic Quality: Identity, and Hedonic Quality: Stimulation (hedonic quality refers to how pleasurable and interesting someone thinks the page is). The users rated the thirty web pages used on a seven point scale along the gambit of each dimension.

The same set of pages were also analyzed by a very structured computer algorithm. Simply put, the algorithm attempts to break the page into pieces that contain an equal amount of information. These quadrants of the image are then measure for balance, symmetry, equilibrium (is the image lopsided or centered), and the total number of quadrants created (an example of the breakdown is shown to the right). The authors used some pretty advanced math, so if you want to check it out for yourself I recommend you read the last page of the article, which is covered in algorithms.

Overall, the algorithm's computation of a page's awesomeness had a strong correlation with what the test users thought as well. Below, you can see how the different measures of attractiveness match up between human and machine. The authors stress the impact of these results- a computer can be used to predict how visually appealing a design is to end-users. This is both a more cost-effective evaluation of a design and a breakthrough in the understanding of how humans make decisions about visuals. Because the testers were only allowed to view the actual page for 150 milliseconds, the results prove that humans make low-level decisions in a similar process as the authors' algorithm does, just in a much more effective way.

Discussion:

This paper killed me. With a title like that, I was not looking forward to reading heaps of procedure and math. However, in the end it turned out to be pretty dang interesting. I wasn't aware that any research was being conducted as to simulating the likes and dislikes of humans based on images and designs. If this work were expanded, then computer programs could screen TV shows, commercials, movies, websites, and photographs to determine if humans would actually find them appealing before launching them publicly or to testers. That's a little bit creepy and a little cool.

The Design of Everyday Things

I wonder if Donald Norman realized that some of the things he wrote about in this book would come to pass by the time we read it, and how he would react to their designs. For instance, his "pocket-size device" that keeps his calendar and reminders mentioned on page 74. He basically describes the calendar functionality of a PDA or an iPhone. Did he know going in that such things were going to exist within twenty years? Did the creators of such things read his book?

Anyway... The Design of Everyday Things focuses on (you guessed it) the design of everyday things. Doors, phones, and VCRs are main characters, with noteworthy performances by projectors, lights, digital watches, and radios in supporting roles. Basically, Norman looks at something we take for granted and breaks it down into its general level of FAIL based on a few main characteristics. He then encourages designers to consider these characteristics in their future work.

Characteristics to keep in mind:

• Constraints (physical, semantic, logical, cultural). Simply put, constraints refine design by limiting it. If you want to make your Lego policeman and bike, you don't put the wheel on the officer's head, make him face backwards, and put the red light on the front of the bike. Constraints are good! Don't fight them.
• Natural Mappings. If you have four stove burners and slap four dials on there to control them, make the layouts match! Otherwise people make mistakes and can't remember stuff. Which leads into...
• Keep Knowledge in the World. If it's natural to use, then people need not focus on memorizing how to operate something. I should be able to pick up a pencil and write without having to first memorize its instruction manual.
• Maintain Visibility. If you want to turn the water on in a sink, don't hide a foot pedal under the counter. Let the user (again, naturally!) see how to use things. Sometimes you must sacrifice elegance for simplicity.
• Give Feedback. If you double-click Internet Explorer and then nothing tells you that it's currently in the process of crashing (no hour glass, warning, etc.) then you waste valuable time and are generally confused. Then you call Geek Squad and ask them what's wrong with your CPU when you don't even know what that means (but you want to sound high-tech). Let users see that what they do has a direct impact on what they're using, and they will feel better about themselves

Now there are some things I didn't mention about designing for people (dummies), but only because I want to make a point of my own. If everything were designed as Norman wishes it to be designed, how would we ever have specialized skills and make advancements? He makes very good points, but if you boil something like this computer down to his ideal, then I wouldn't be able to get online and navigate to this website without Microsoft's Clippy by my side. Very frustrating! Some people don't need such a basic and natural design. Sometimes a little bit of elegance and required level of skill is a good thing. Boosh!