Identifying user-dependent information that can be automatically collected helps build a user model by which 1) to predict what the user wants to do next and 2) to do relevant preprocessing. Such information is often relational and is best represented by a set of directed graphs. A machine learning technique called graph-based induction (GBI) efficiently extracts regularities from such data, based on which a user-adaptive interface is built that can predict next command, generate scripts and prefetch les in a multi task environment. The heart of GBI is pairwise chunking. The paper shows how this simple mechanism applies to the top down induction of decision trees for nested attribute representation as well as nding frequently occurring patterns in a graph. The results clearly shows that the dependency analysis of computational processes activated by the user commands which is made possible by GBI is indeed useful to build a behavior model and increase prediction accuracy.

Reviewing past events has been useful in many domains. Videotapes and flight data recorders provide invaluable technological help to sports coaches or aviation engineers. Similarly, providing learners with a readable recording of their actions may help them monitor their behavior, reflect on their progress, and experiment with revisions of their experiences. It may also facilitate active collaboration among dispersed learning communities. Learning histories can help students and professionals make more effective use of digital library searching, word processing tasks, computer-assisted design tools, electronic performance support systems, and web navigation. This paper describes the design space and discusses the challenges of implementing learning histories. It presents guidelines for creating effective implementations, and the design tradeoffs between sparse and dense history records. The paper also presents a first implementation of learning histories for a simulation-based engineer...

We describe a mixed-initiative framework designed to support the customization of complex graphical user interfaces. The framework uses an innovative form of online GOMS analysis to provide the user with tailored customization suggestions aimed at maximizing the user’s performance with the interface. The suggestions are presented non-intrusively, minimizing disruption and allowing the user to maintain full control. The framework has been applied to a general userproductivity application. A formal user evaluation of the system provides encouraging evidence that this mixedinitiative approach is preferred to a purely adaptable alternative and that the system’s suggestions help improve task performance.

This paper analyzes trends in the approach to evaluation taken by CHI papers in the last 24 years. A set of papers was analyzed according to our schema for classifying type of evaluation. Our analysis traces papers ’ trend in type and scope of evaluation. Findings include an increase in the proportion of papers that include evaluation, and a decrease in the median number of subjects in quantitative studies. We also critique the types of subjects, in particular an over reliance on students, and lack of appropriately gender balanced samples. We contextualize these findings in historical trends as we move from machines intended for the technical elite in laboratories to computers integrated into the daily life of everyone.

We introduce GEKA, a graphically enhanced keyboard accelerator method that provides the advantages of a traditional command line interface within a GUI environment, thus avoiding the “Fitts-induced bottleneck ” of pointer movement that is characteristic of most WIMP methods. Our design rationale and prototype development were derived from a small formative user study, which suggested that advanced users would like alternatives to WIMP methods in GUIs. The results of a controlled experiment show that GEKA performs well, is faster than menu selection, and is strongly preferred over all mouse-based WIMP methods.

Identifying user-dependent information that can be automatically collected helps build a user model by which to predict what the user wants to do next. Such information is often relational and is not suited to the traditional attribute-value representation. Graph-based induction (GBI), a machine learning technique to find typical patterns in a directed graph, efficiently extracts rules to predict next command from such data. The heart of GBI is pairwise chunking. The paper shows how this simple mechanism applies to the top down induction of decision trees for nested attribute representation. The algorithm is implemented and tested against both artificial and real data. The results clearly shows that the dependency analysis of computational processes activated by the user commands which is made possible by GBI is indeed useful to build a behavior model and increase prediction accuracy. 1 Introduction Finding regularities in data is a basis of knowledge acquisition, and extracting beha...

Many modern human-computer interfaces are difficult for people to use. This is often because these interfaces make no significant attempt to communicate with the people who use them. In other words, these interfaces are uncooperative: They do not adapt themselves to their users' needs and they are insensitive to human foibles. Ordinary command line interfaces such as that of the UNIX C shell (csh) are intolerant of even the most simple input errors, even when those errors have obvious corrections. An "intelligent" UNIX shell interface, on the other hand, would make use of knowledge and interaction context in order to interpret --- and as necessary, correct --- its users' commands. Valet is a prototype of such an "intelligent" interface to the UNIX C shell. Valet adds knowledge-based parsing and input correction to the shell by encapsulating an ordinary C shell process within a framework that allows Valet to control the shell's input and output. Valet intercepts shell commands and par...

History-keeping has surfaced as a potentially valuable asset to educational and other software. Current research in learning histories considers the hypothesis that providing learners with a readable record of their actions may help them monitor their behavior and reflect on their progress. However, the scope of learning histories goes far beyond the means provided by an undo/redo or document-recall history system. In this paper we describe Trails, a component-based framework for constructing rich learning history modules based on the client/server model. Trails historians are loosely-coupled to their client applications and interact with them through a set of welldefined interfaces. Trail historians also provide ample means for history visualization and direct manipulation. The client-server architecture facilitates history extensions to existing applications, while the modular design promotes experimentation with different visualization metaphors.