There are a wide range of projects which apply AI techniques to create advanced interfaces - and which use the challenges of human interaction to drive the development of novel AI and machine learning techniques. Many projects are listed on this page, but see also the incredible innovations at the Ubicomp Lab.
Accessible Goal Crossing Project
The Accessible Goal Crossing Project is funded by the National Science Foundation as an effort to create user interfaces more suited to people with motor impairments by using goal crossing instead of pointing-and-clicking, which is difficult for many people. Although goal crossing has been investigated on pen and tablet computers for use by able-bodied users, it presents substantially different challenges when used on the desktop with cursor control devices by people with motor control problems.
Crowdsourcing Taxonomy Creation
Our goal is to break down large projects into work that can be done in parallel microtasks. Parallel crowd algorithms enable hundreds of people to contribute at once. Breaking tasks into short (typically less than 1 minute) micro tasks makes it more attractive for people to make a contribution in their free time. A major application for crowd algorithm is using people to organize data. For example, given 100 travel tips, we can create a taxonomy that shows what the popular topics are, what topics are missing, and let's users navigate the data more easily.
Decision-Theoretic Control for Crowdsourced Workflows
Crowd-sourcing is a recent framework in which human intelligence tasks are outsourced to a crowd of unknown people (”workers”) as an open call (e.g., on Amazon’s Mechanical Turk). Crowd-sourcing has become immensely popular with hoards of employers (”requesters”), who use it to solve a wide variety of jobs, such as dictation transcription, content screening, etc. In order to achieve quality results, requesters often subdivide a large task into a chain of bite-sized sub-tasks that are combined into a complex, iterative workflow in which workers check and improve each other’s results.
Prefab
Prefab is a system for implementing advanced behaviors through the reverse engineering of the pixels of graphical interfaces. In other words, Prefab explores the question: what if every GUI was open source? One example of what Prefab enables is better HCI research: Prefab can be used by HCI researchers to test many of their ideas using existing applications, rather than implementing a toy prototype within an isolated testbed.
SUPPLE: Decision-Theoretic UI Generation
SUPPLE is an application and device-independent system, developed at University of Washington, that automatically generates user interfaces for a wide variety of display devices. SUPPLE uses decision-theoretic optimization to render an interface from an abstract functional specification and an interchangeable device model.
Utility: Measure What Matters
A system is useless if it can’t motivate people to use it. Your interfaces must be clear and appealing, your tasks must be satisfying and rewarding, and your games must be fun.
Here is a way to evaluate a system’s ability to recruit use, or human attention. Let’s take any web interface and task. Now let’s post it as a job on Mechanical Turk and see how much we need to pay people to use the interface to complete the task. The less we have to pay workers, the better the interface and task.
Of Historical Interest
Intelligent Wikipedia
Berners-Lee's compelling vision of a Semantic Web is hindered by a chicken-and-egg problem, which can be best solved by a bootstrapping method, creating enough structured data to motivate the development of applications. However, automatic information extraction systems produce errors and are not tolerated by users, whereas user contributions incentives and management to control vandalism.
reform
There is a lot of useful information on the Internet, but webmasters do not always present it in the best way. Reform lets end users put a new face on webpages, without subjecting them to the whims of a webmaster, and without learning to program themselves.