IUI 09 - Do You Know?

Do You Know is a friend recommendation widget developed by the authors for use in IBM's internal social networking system. The authors developed DYK with a novel UI which they proposed would help make accurate and welcome connections. One of the major features of DYK is the extensive list of reasons why the person was recommended. They interviewed users after the testing period and one user response stuck out to me. They thought it should be more clear about what happens if you click 'No Thank You'. They might not click 'No' for someone if that person would be notified.

Thoughts

I think recommendation systems further separate our social worlds. There is already a sense of real life friends vs online friends. I think that is should be real life connections that start friendships. If you have enough in common with someone that they are recommended by software then you probably already know them and if you are not online friends already then theres probably a good reason.

CHI 09- Lightweight Tagging

The paper explores the use of tagging in the setting of a personal computer. Using tagging to support new ways of working with and accessing applications and documents. They created an application which runs within Windows that supports the generic tagging of applications and documents. They collected data across a group of users who used the system on their work computers to help with their jobs. They found that the tags were particularly useful when working with information from diverse data stores and that users were able to create new types of data organizations which bridge folder systems and document types.

Thoughts

The idea of being able to tag items in my work computer seems useful. Maybe this same concept could be used in a IDE to tag files or specific methods. Then you could easily pull up a small specific working set of code.

CHI 09- EyeSpy

The authors created a game to encourage the creation of an accurate landmark image database.

The game consisted of taking pictures of landmarks and accurately describing them. For each player that went to and confirmed your picture of the landmark you received points. You would also receive points for locating and verifying the pictures other players had taken. The goal was create a large database of accurately described and photographed landmarks which would be used in a navigation application.

Thoughts

The game doesn't seem all that fun to me. It would be better if the descriptions were riddles or something but that wouldn't work at all.

CHI 09- Dynamically Changeable Buttons

Chris Harrison, Scott E. Hudson - Carnegie Mellon University, Pittsburgh, PA, USA

The authors attempt to combine the low-attention and vision-free interactions possible with physical buttons with the flexibility of a touchscreen display. They create an interface layer that consists of a deformable surface on top of a display screen. For their user study they measured the amount of glances required to make a button touch using four different interfaces, a flat surface, physical buttons, their system with the button sticking out, and sunken in. The results showed that their system in both configurations required less glances than the physical buttons or a flat surface.

Thoughts

Physical buttons are irreplaceable in my opinion. I can't imagine ever giving up my plastic button keyboard for touch pad version. I liken their system to the buttons on some cheap remote controls, a sheet of plastic with bumps for buttons. The problem was never finding the buttons but getting it to register clicks. Anyway, the methods of created physical buttons on top of a display is a great idea however it will never be widely used if it uses a projector and camera to sense input because the setup is just too large.

CHI 09- Remembrance of things tagged

Raluca Budiu - Nielsen Norman Group, Fremont, CA, USA

Peter Pirolli, Lichange Hong - Palo Alto Research Center, Palo Alto, CA, USA

This paper sought to find the effects of tagging system on tag production and information retention. They compared two different tagging systems, one you simply had to select a word or set of words and then click Ok and the other required the user to type the tags into a text field. Using these 2 systems and the base case of no tagging system the authors wanted to see how each system affected the amount of tags created, the users' reading speed, fact recall and recognition. The results showed that tagging did not significantly increase recall or recognition.

Thoughts

It is interesting to see a study about how tagging methods and tagging in general affect our ability to read and remember information. I thought for sure that tagging would significantly increase the recall and recognition rates and that type-to-tag would do so more than click-to-tag. Maybe since tags reduces information to only its keywords we somehow forget it easier.

CHI 09- Butler Lies

Jeff Hancock, Jeremy Birnholtz, Natalya Bazarova, Jamie Guillory, Josh Perlin, Barrett Amos

Department of Communication and Information Science Program

Cornell University

Ithaca, NY 14853

The authors introduce "butler lies" which they define as an act deception to manage social interactions. Butler lies are used to avoid and exit conversations, and to explain communication behavior. Examples would be like saying, IMing "cant talk, im studying" when you are not studying, just don't want to talk to that person. Or maybe telling someone "sorry I missed your call I was in the shower" when really you just didn't want to talk to them then. These lies are rampant in our society because of our many forms of 'always-on' communication such as cell phones, text messaging and email.

Thoughts

I think its awesome that someones finally coined a term to cover these little lies. 'Butler lie' has a fitting connotation in my opinion like 'respectful, nothing-personal, lie'. It is just like if someone came to your house and knocked, and you told your butler to tell them you weren't home. In fact that's probably why they chose butler lie. I think that the definition of a butler lie should also include deception to start a new conversation, such as, 'oh sorry my phone must have called you on accident, but anyway whats up?' or sending a 'i had a great time last night' text message to ex girlfriend by accident just to make her jealous.

CHI 09- Social TV

David Geerts and Dirk De Grooff IBBT / K.U.Leuven, Leuven, Belgium

Social television enables remote communication and interaction between viewers. Many social TV systems are currently in development. There are existing guidelines for ensuring a usable system however there are no guidelines evaluating the social aspects of a system. The authors propose a set of twelve heuristics for measuring the sociability of these TV applications.

Their heuristics in my own words:

1. Offer text and voice communication.
2. Have a user status bar with options like busy, away, watching intently, watching relaxed, and not really watching at all just sitting on the couch with the TV on.
3. Let people collaborate at the while watching at the same time or if they watch at different times.
4. Allow for interactions between people at different TV and also people watching at the same TV.
5. Track and use everyone's TV choices and habits
6. Let the user change the settings.
7. Have privacy options and ensure privacy.
8. Don't distract from the actual TV watching.
9. Be sure and tell user about changes and upcoming shows or events.
10. Recognize the genres of programs and behave appropriately for each genre. Like don't allow for people to spoil the endings of mystery shows but for comedy shows it doesn't really matter.
11. Allow for sharing between devices like a PC or smartphone.
12. Encourage shared activities.

The major effort of the authors was spent on grounding these sociability heuristics in observed user behavior. They have not validated these heuristics yet because no social TV systems are commercially available.

Thoughts
The heuristics they propose seem like 'duh' to me. I am assuming that these authors are attempting to create a social TV system and one of their first steps to research the design requirements. After they gathered the requirements they thought 'Hey we could publish a paper about these' and that's just what they did. All the heuristics are good ideas in my opinion and I agree that a system that fulfilled them all would be a success.