Martin Dostál

Fine-Grained Adaptive User Interface for Personalization of a Word Processor: Principles and a Preliminary Study

We introduce novel fine-grained adaptive user interface for the OpenOffice.org Writer word processor. It provides a panel container with personalized user interface to an individual user. The personalization is provided on frequently or recently used user commands, user’s preferred interaction styles and frequently used parameters to user commands. The next part of the paper describes a proof-of-concept usability study. We measured task completion times and error rates on the proposed adaptive interface, menus and toolbars on a word processing task.

Multimodal navigation display

In this paper we discuss the opportunities of multimodal interaction for multimodal flight planning in a navigation display and introduce key design principles of multimodal interaction (MMI) via keyboard, cursor-control device, touch gestures and voice. We present results of human factors evaluation of multimodal flight planning conducted on an interactive prototype and outline the possibilities of multimodal interaction for other functions in the flightdeck.

On the Differences in Usage of Word Processing Applications

An analysis was conducted to evaluate the differences between users in utilization of functionality and interaction styles in word processing applications. The study evaluated 117 users of various word processing applications using a questionnaire. Results show that differences between users in used functionality are considerable. The differences in utilization of interaction styles on a common functionality was found substantial. In addition, we show that the differences in utilization of functionality do not correlate much with the differences in utilization of interaction styles, i.e., users which has similar usage patterns in utilization of functionality do not have much similar usage patters in utilization of interaction styles.

A research framework for performing user studies and rapid prototyping of intelligent user interfaces under the OpenOffice.org suite

We introduce a research framework which enables to use the OpenOffice.org as a platform for HCI research, particularly for performing user studies or prototyping and evaluating intelligent user interfaces. We make two contributions: (1) we introduce an innovative hybrid logging technique which provides high-level, rich and accurate information about issued user commands, command parameters and used interaction styles. Our logging technique also avoids an unwanted requirement for further complex processing of logged user interface events to infer user commands, which must be performed on most loggers these days. (2) Our logging tool acts as a component object in OpenOffice.org with an easy-to-use Application Program Interface (API) which enables, along with deep OpenOffice.org programmability, to use OpenOffice.org as a research framework for developing and evaluating intelligent user interfaces.

Adaptive User Interface Personalization in ERP Systems

Enterprise Resource Planning (ERP) systems had grown into complex applications integrating various processes and large amounts of data in detail. Such complexity negatively impacts usability owing to providing single static user interface that should meet users with diverse roles, tasks and requirements. In this paper we introduce a concept of adaptive user interface personalization in the SAP ERP system. We address two important aspects of personalization; (1) personalization to the user’s prominent tasks, (2) personalization of individual on-screen task forms using adaptive progressive disclosure with ephemeral visualization. In the next we describe a mock-up of our adaptive system in detail and results of brief questionnaire study.

An Analysis of Usage Patterns in Utilization of Interaction Styles

This paper introduces particular results on analysis of usage patterns in contemporary word processing applications. We evaluate how interaction styles are used accordingly to frequency of usage of underlying user commands or semantical categories of user commands. We show that menus are used as a prevalent interaction style and report usage patterns on toolbars and keystrokes. Furthermore, the differences in usage of toolbars and keystrokes on frequently used commands are evaluated.

A Hybrid Approach to User Activity Instrumentation in Software Applications

This paper introduces novel approach to logging user activity in software applications that provide accurate high-level information about issued commands, commands parameters and user interaction styles without a need for inferring user activity from logged user interfaces events as it is required on most loggers these days. We also demonstrate the proposed approach on a prototype implementation under the OpenOffice.org suite.