In this course, you will learn relevant fundamentals of human motor performance, perception, and cognition that inform effective interaction design. You will use these models of how people work to design more effective input and interaction techniques. You’ll apply these to both traditional graphic and gestural interfaces.
Character Study
Loading...
來自 University of California San Diego 的課程
输入与互动
534 個評分
從本節課中
Input
There’s more to interfaces than what’s on the screen. While often overlooked, input deserves to be on equal footing with its more popular sibling, output. Also, input is a rare case where we can model user behavior mathematically. In this module, you’ll learn how good input is more than just preference, trace input from the fingertip to the screen, and think about the diversity of possible input devices and their relative merits.
與講師見面
Scott KlemmerProfessor
Cognitive Science & Computer Science
Descartes famously said I think, therefore I am.
And in doing so, it advocated a view that the mind and the body are wholly distinct.
And this Cartesian dualism often shows up when we think about interacting with
computers.
People often tell me that they wish they had an interface
that provided a direct interaction with their brain,
and the intuition here is that your brain has lots of ideas
that somehow we just can't get out through our bodies.
And it's absolutely the case that we have lots of thoughts that
are hard to get out or act out or express.
But I want to push back a little bit on the intuition that the problem
is the bodily channels for passing those thoughts through.
Our brains and our bodies evolve together.
The organization and function of the brain
evolved in response to our bodily capacities and vice versa.
What this means is it's not like there's a massive chunk of cortex somewhere in here.
That's got no output,
like a computer with an empty port at the back just waiting to be connected.
There'd be no reason to evolve a brain that had all this empty cortex.
And further more, keeping the brain functioning is extremely energy intensive.
To have an unused part of the brain would be a total waste which your welcome to
remind your teenagers anytime their watching television.
This mind body integration runs really deep.
Here's an image of the homunculus.
It illustrates the sensory motor cortex and
how much cortical real estate is devoted to different parts of the body.
One thing that really jumps out is the amount of cortical real estate
that's devoted to our hands.
That's a really high bandwidth channel.
Our mind body integration doesn't mean that there's no room for improvement, or
that all input technologies are equal.
Let's look at a powerful but mundane example, the keyboard.
In English, the top row of the keys along the keyboard spells QWERTY.
As many of you probably know, the QWERTY keyboard
design emerged in response to a mechanical challenge with a typewriter,
You've got the hammers that are attached to each of the keys so that when you press
on a key that's flipping a hammer, which has a letter on the end of it,
that's gonna hit the sheet of paper that you're typing on.
If the user types adjacent keys on a typewriter,
that springs adjacent hammers into action.
And while one's going down and the other's going back, they can jam.
This QWERTY typewriter
layout was invented in the early 1870s by Christopher Latham Sholes.
It solved the hammer collision problem by drawing on a wonderful combination
of computational analysis and iterative design with users.
In English, and in all languages, some characters appear more often than others.
In English, the letter forms earning
the most air time are e, T, A, I, O, N, S.
E is the most common letter followed by these.
Furthermore, some pairs of letters called bigrams, appear more commonly than others.
Shaw's collaborator, Enos Densmore,
tallied the bigram use in the King James Bible, and
found the following letter pairs to be especially common.
Shows innovation, like somebody seeking to break up a bar fight.
Was to take these letter pairs that co-occur, put one on one side of
the typewriter keyboard, and the other on the other side of the typewriter keyboard.
You may have heard that the QWERTY keyboard design is not very efficient.
After all, it was created to solve this mechanical problem that no longer exists.
Economists describe technologies and decisions whose design
reflects historical constraints rather then current reality as path dependence.
I think that the last time I used a mechanical typewriter was filling
out a graduate school application form.
And while jamming hammers would make a great funk band name for
fast typists, it's not a major problem in my life.
So we have a layout that was designed expressly to combat
the problem furthermore,
we have a layout that was designed expressly to combat the problem of speed.
People type quickly, the hammers jam, and you got a problem.
So this can't be a good design, right?
Actually, no.
Physically separating the hammers to remove mechanical collisions also,
temporally separated the hand rolls to reduce biomechanical collisions.
What I mean is this, as I'm typing, like this, one hand is going down hitting
the first key and the other hand can be preparing to hit the next key.
And this alternation, one down while the other goes up,
is actually an extremely efficient design.
The QWERTY keyboard is just one example of data driven layout
based on usage frequency.
Lead type printers for English newspapers favorite
a frequency based arrangement this e, T, A, I, O, N,S.
And when they made a mistake, type setters for the newspaper,
they would run their fingers across the row that began with s.
In fact that row is S, H, R, D, L, U.
Run your finger across this a few times
as a warning that there's a typo that needs to be fixed.
But of course the warning didn't always get pulled out of the lead that came out
of the machine.
Every once in a while, this would actually end up in the newspaper.
It's like the lorem ipsum that shows up on so many webpages of the analogue age.
Mad Magazine thought this was funny, and so
they would put it in their magazine from time to time.
And the famous computer scientist, Terry Winograd took up this joke as the name of
his seminal natural language understanding system, which he called SHRDLU.
These first two examples both come from English, but of course,
they also apply to other character sets.
One of the most interesting examples is Chinese, with its myriad logograms.
Here's an example of a Chinese typewriter that has
more than 25 characters placed within it.
When these characters were type set by people, especially, for
efficiency purposes, the most frequent letter forms were placed in
the middle of the job tray, the big wooden box that contains all the characters.
And what the historian, Thomas Mullaney, pointed out to me is
that then you can use the layout of the job tray of all of the characters as
a means of understanding the frequency of character use in the language.
And one really interesting example of this is if you look at job trays from
the early 20th century in China there character Mao was present in
the periphery of the tray if at all.
Of course 50 or
60 years later the character was much more central to printing in China.
Today we described the system used by all of these examples as one of optimization.
None of these keyboards have the benefit of computers or
formal optimization algorithms to decide the best layout.
The data was tallied manually and iteration was trial and error.
If we did have the benefit of modern computation how much better could we do?
Schuman and Ji colleagues have spent a decade investigating
computationally designed keyboards.
It draws on motor models like Fitz law
that you learned about in the earlier videos.
If we know from Fitz law how long it takes to acquire a target here,
how long it takes to press on a key.
And we know from language models how frequent those different characters and
character combinations are, then it's in principle possible to design a key board
that minimizes the total amount of travel that a typist must engage in.
And this is exactly what Schuman and his colleagues did.
Now for many of us desktop keyboards are fast enough
most of the time for practical purposes.
It's clearly possible to find a keyboard that's slightly faster.
In fact, several designs exist, you may use one, but the benefits
are relatively modest and desktop typing speeds are generally sufficient.
For a good touch typist, the barriers is usually coming up with stuff worth typing.
By contrast, input on mobile devices is a heck of a lot slower, even for
your 17-year-old who's got the Olympic rate thumb movement.
So if you want to make a big impact in user interface design and
convince people to switch.
It's best to look for a place where there's currently a lot of pain.
Consequently the literature has hundreds of examples of different mobile keyboard
designs, and I'd like to talk about Schuman's work today.
Because I think it's the next insight
after just the computational layout that the really transformative one.
Because a better layout key can significantly improve layout on a mobile
device but embarrassingly, we're kind of entrenched and used to what we have.
Also, even if we move the right characters next to each other in the right ways,
in a character-by-character entry system, I still need to put
a digit down on one key, raise it up, and put it down on another key.
Or, in a two-thumb or multi-finger system, I can oscillate back and forth.
What if you didn't have all of that up-and-down,
that would really speed things up significantly.
So that's what Schumann did, is if you were able to put your finger down at
the beginning of the word, or stylus down at the beginning of a word, go through
all of the letters that you might touch and then release at the end of a word.
That's what Sherman's ShapeWriter system did.
And this word based entry system opened the door for many subsequent innovations.
You've probably even tried some of these.
And I think it's a really example, interesting example of how,
especially once your input devices computationally rendered at every frame.
And you have lots of smarts underneath the hood.
There's some really radical transformations of input devices
that are possible.
The last example that I wanna show today comes from Jacob Wobbrock and
his colleagues.
Desktop keyboards and the QWERTY design are relatively entrenched in English.
In the mobile world we have some more flexibility for
something like a word based input system to really come and take things over.
Another group of users for whom input is extremely difficult and
slow are users who have motor impairments.
In fact, all of us have motor limitations.
And some more so than others and of different sorts, and so
there's a big opportunity there.
EdgeWrite used a stylized alphabet system, similar to what,
there was a system called Graffiti for the Palm Pilot.
And the innovation of EdgeWrite It was a piece of plexiglass
that provided a square hole through which you would enter with a stylus or finger.
Now, what's nice about this, is it offers a physical constraint.
So if I have to draw a straight line, I'm not very good.
That's one reason we use a ruler when we need to.
The edge right system gives you that physical constraint.
And what that means, especially for users with mobility limitations,
is you can lean against the sides and enter a lot more quickly.
What we've seen with all of these designs in the video,
is that thinking about what people are gonna type.
And who's gonna be doing the typing and what the setting for
the typing is, can give you new innovations in how input design
layout works to create a much more fluid, fast, efficient experience.
